Datasets:

james-burton

/

aug-text-exps-v3

like 0

SVC

C1

Health Care Services Satisfaction Prediction

The prediction probability associated with class C2 and class C1, respectively, is 35.34% and 64.66%. Based on these probabilities, the model labels the given case as C1 since it is the most probable class. According to the attribution analysis, the most relevant features considered by the model here are F16, F7, and F6, while the least relevant features are F15, F5, and F12. Regarding the direction of influence of the features, F16, F7, F6, and F3 are the top positively supporting features, driving the decision higher in favour of C1. Further increasing the probability that C1 is the true label are the values of other positive features such as F13, F9, F2, and F4. To explain why the likelihood of C2 is 35.34%, we have to look at the negative contributions from F14, F1, F11, F5, F15, and F12. The abovementioned negative features contradict the model's decision with respect to the classification outcome.

{'C2': '35.34%', 'C1': '64.66%'}

{'F16': 'waiting rooms', 'F6': 'Hygiene and cleaning', 'F7': 'Specialists avaliable', 'F14': 'Quality\\/experience dr.', 'F3': 'Modern equipment', 'F1': 'Exact diagnosis', 'F13': 'hospital rooms quality', 'F2': 'Check up appointment', 'F11': 'avaliablity of drugs', 'F9': 'friendly health care workers', 'F4': 'Time waiting', 'F10': 'Communication with dr', 'F8': 'lab services', 'F5': 'parking, playing rooms, caffes', 'F15': 'Time of appointment', 'F12': 'Admin procedures'}

{'F14': 'F16', 'F4': 'F6', 'F7': 'F7', 'F6': 'F14', 'F10': 'F3', 'F9': 'F1', 'F15': 'F13', 'F1': 'F2', 'F13': 'F11', 'F11': 'F9', 'F2': 'F4', 'F8': 'F10', 'F12': 'F8', 'F16': 'F5', 'F5': 'F15', 'F3': 'F12'}

{'C2': 'C2', 'C1': 'C1'}

Satisfied

{'C2': 'Dissatisfied', 'C1': 'Satisfied'}

KNeighborsClassifier

C1

Real Estate Investment

The classifier is very uncertain about the correct label for the case given. Regarding the classifier's decision, there is close to an even split on the probability of either of the possible labels is the correct label but the classifier chooses the label as C1. The prediction verdict above is attributed to the contributions of mainly the following features: F2, F14, F13, and F10, however, the lowest ranked features are F4, F1, and F15. Analysing the direction of influence of the features shows that there are ten positive and ten negative features. Positive features such as F13, F10, F19, and F6 increase the response of the classifier in favour of the assigned label. Conversely, negative features such as F2, F14, F16, and F11 decrease the likelihood of C1 being the correct label given that their values support the alternative label, C2. The uncertainty concerning the label assignment can be due to the fact that the top negative features F2 and F14 have very high attributions shifting the classifier's verdict away from the C1 class.

{'C1': '50.00%', 'C2': '50.00%'}

{'F2': 'Feature7', 'F14': 'Feature4', 'F13': 'Feature2', 'F10': 'Feature8', 'F19': 'Feature20', 'F16': 'Feature1', 'F11': 'Feature12', 'F6': 'Feature15', 'F20': 'Feature6', 'F7': 'Feature9', 'F8': 'Feature17', 'F12': 'Feature3', 'F17': 'Feature19', 'F9': 'Feature13', 'F3': 'Feature18', 'F18': 'Feature5', 'F5': 'Feature11', 'F4': 'Feature16', 'F1': 'Feature10', 'F15': 'Feature14'}

{'F11': 'F2', 'F9': 'F14', 'F1': 'F13', 'F3': 'F10', 'F20': 'F19', 'F7': 'F16', 'F15': 'F11', 'F4': 'F6', 'F10': 'F20', 'F12': 'F7', 'F6': 'F8', 'F8': 'F12', 'F5': 'F17', 'F16': 'F9', 'F19': 'F3', 'F2': 'F18', 'F14': 'F5', 'F18': 'F4', 'F13': 'F1', 'F17': 'F15'}

{'C2': 'C1', 'C1': 'C2'}

Ignore

{'C1': 'Ignore', 'C2': 'Invest'}

DecisionTreeClassifier

C1

Car Acceptability Valuation

The classification algorithm believes that C1 is the output label that was generated with 100% certainty and that C2 is unlikely to be the correct label in this case. According to the attribution investigations, the following input features are ranked from most relevant to least relevant: F6, F3, F5, F2, F1, and F4. As shown by the attribution plot, F6 is the only one shown to positively contribute to the above classification decision, while the others contribute negatively. The contributions of negative features such as F3, F5, and F2 result in the decision being driven in a different direction. From the prediction confidence level, we can conclude that the very strong influence of F6 overshadows the contributions of the negative features hence the very high confidence level.

{'C1': '100.00%', 'C2': '0.00%'}

{'F6': 'safety', 'F3': 'persons', 'F5': 'buying', 'F2': 'maint', 'F1': 'lug_boot', 'F4': 'doors'}

{'F6': 'F6', 'F4': 'F3', 'F1': 'F5', 'F2': 'F2', 'F5': 'F1', 'F3': 'F4'}

{'C2': 'C1', 'C1': 'C2'}

Unacceptable

{'C1': 'Unacceptable', 'C2': 'Acceptable'}

RandomForestClassifier

C3

Flight Price-Range Classification

Of the three possible labels, there is 100.0% confidence that C3 is the most probable label for the given case. The features that heavily influence the classification verdict presented here are F10, F5, and F2, and they have a very strong positive contribution, increasing the odds of the C3 prediction. Other features with a positive influence on the model are F6, F7, F4, F3, and F8. On the contrary, F9, F11, and F12 make the model's decision fluctuate negatively towards selecting an alternative label. All of the negative features mentioned above have a low to moderate impact on the classification verdict presented here compared to F2, F5, and F10. Finally, F1 with its very low positive impact is the least ranked feature marginally pushing the decision towards the assigned label.

{'C3': '100.00%', 'C1': '0.00%', 'C2': '0.00%'}

{'F2': 'Duration_hours', 'F10': 'Airline', 'F5': 'Total_Stops', 'F7': 'Journey_day', 'F9': 'Source', 'F6': 'Destination', 'F8': 'Journey_month', 'F3': 'Dep_minute', 'F4': 'Arrival_minute', 'F11': 'Arrival_hour', 'F12': 'Duration_mins', 'F1': 'Dep_hour'}

{'F7': 'F2', 'F9': 'F10', 'F12': 'F5', 'F1': 'F7', 'F10': 'F9', 'F11': 'F6', 'F2': 'F8', 'F4': 'F3', 'F6': 'F4', 'F5': 'F11', 'F8': 'F12', 'F3': 'F1'}

{'C3': 'C3', 'C2': 'C1', 'C1': 'C2'}

Low

{'C3': 'Low', 'C1': 'Moderate', 'C2': 'High'}

RandomForestClassifier

C2

Used Cars Price-Range Prediction

Per the model, class C1 has a prediction probability of 10.50 percent, whereas class C2 has a predicted probability of 89.50 percent. As a result of the model, it can be determined that C2 is the most likely label for the given scenario. All of the input features are shown to contribute to the above conclusion, with F2, F6, and F8 having the most influence on the classification decision. The least influential features with regard to this classification are F4, F9, F7, and F10, whereas, the impact of F5, F1, and F3 can be classified as modest. The large positive contributions of F6 and F2 are responsible for the model's high confidence which further supported by the positive contributions of F5, F4, and F9. In conclusion, the negative features F8, F1, F7, F3, and F10 favour labelling the case as C1 hence the associated predicted probability.

{'C1': '10.50%', 'C2': '89.50%'}

{'F6': 'Power', 'F2': 'car_age', 'F8': 'Transmission', 'F5': 'Fuel_Type', 'F1': 'Name', 'F3': 'Mileage', 'F4': 'Engine', 'F7': 'Owner_Type', 'F9': 'Kilometers_Driven', 'F10': 'Seats'}

{'F4': 'F6', 'F5': 'F2', 'F8': 'F8', 'F7': 'F5', 'F6': 'F1', 'F2': 'F3', 'F3': 'F4', 'F9': 'F7', 'F1': 'F9', 'F10': 'F10'}

{'C2': 'C1', 'C1': 'C2'}

High

{'C1': 'Low', 'C2': 'High'}

AdaBoostClassifier

C2

Basketball Players Career Length Prediction

The classifier says that C2 is the most likely label for the provided data with relatively high confidence. It is crucial to remember, however, that there is a 21.80% possibility that it is C1. F12 and F10 are the major driving variables for the aforementioned classification or prediction choice. The remaining variables F5, F13, F19, and F3 have a modest to minor impact on the selection made above. Among the input variables, F5, F3, F4, F6, and F11 are the subset that have a negative influence or contribution whereas all of the remaining variables have a positive impact. In essence, the substantial positive contributions of F12 and F10, together with the contributions of additional positive variables such as F13, F19, F18, and F7, account for the classifier's confidence in this classification.

{'C2': '78.20%', 'C1': '21.80%'}

{'F12': 'GamesPlayed', 'F10': 'PointsPerGame', 'F5': 'Steals', 'F13': 'MinutesPlayed', 'F19': 'DefensiveRebounds', 'F3': 'Rebounds', 'F18': 'Blocks', 'F7': 'FreeThrowAttempt', 'F4': 'FieldGoalPercent', 'F14': 'FreeThrowMade', 'F6': 'OffensiveRebounds', 'F1': 'FieldGoalsMade', 'F9': '3PointAttempt', 'F16': 'FreeThrowPercent', 'F15': '3PointMade', 'F17': 'FieldGoalsAttempt', 'F2': 'Turnovers', 'F11': 'Assists', 'F8': '3PointPercent'}

{'F1': 'F12', 'F3': 'F10', 'F17': 'F5', 'F2': 'F13', 'F14': 'F19', 'F15': 'F3', 'F18': 'F18', 'F11': 'F7', 'F6': 'F4', 'F10': 'F14', 'F13': 'F6', 'F4': 'F1', 'F8': 'F9', 'F12': 'F16', 'F7': 'F15', 'F5': 'F17', 'F19': 'F2', 'F16': 'F11', 'F9': 'F8'}

{'C2': 'C2', 'C1': 'C1'}

More than 5

{'C2': 'More than 5', 'C1': 'Less than 5'}

RandomForestClassifier

C2

Printer Sales

C2 has an 83.0% chance of being the correct label for the case under consideration, making C1 the least likely class with a predicted likelihood of 17.0%. F17, F21, and F24 features have a significant impact on class selection here while on the other hand, the remaining features are shown to have marginal to no contribution to the classification verdict here. In actual fact, the values for F13, F14, F12, F22, F3, and F20 may have been ignored by the classifier because their respective influences are almost zero. Of the important features, only F8, F1, F7, F26, F25, and F6 are negative and this is mainly because their contribution to selection tends to reduce the chance that C2 is the correct label, preferring that the case is classified as C1. The remaining features such as F17, F21, F24, F18, F9, and F11 strongly contribute positively, increasing the chances of C2 which explains the level of certainty associated with C2.

{'C2': '83.00%', 'C1': '17.00%'}

{'F17': 'X8', 'F21': 'X24', 'F24': 'X1', 'F9': 'X2', 'F18': 'X10', 'F11': 'X15', 'F8': 'X25', 'F23': 'X23', 'F2': 'X18', 'F1': 'X4', 'F7': 'X7', 'F19': 'X17', 'F10': 'X3', 'F4': 'X22', 'F26': 'X5', 'F25': 'X9', 'F6': 'X12', 'F15': 'X19', 'F5': 'X11', 'F16': 'X16', 'F13': 'X14', 'F14': 'X21', 'F12': 'X20', 'F22': 'X13', 'F3': 'X6', 'F20': 'X26'}

{'F8': 'F17', 'F24': 'F21', 'F1': 'F24', 'F2': 'F9', 'F10': 'F18', 'F15': 'F11', 'F25': 'F8', 'F23': 'F23', 'F18': 'F2', 'F4': 'F1', 'F7': 'F7', 'F17': 'F19', 'F3': 'F10', 'F22': 'F4', 'F5': 'F26', 'F9': 'F25', 'F12': 'F6', 'F19': 'F15', 'F11': 'F5', 'F16': 'F16', 'F14': 'F13', 'F21': 'F14', 'F20': 'F12', 'F13': 'F22', 'F6': 'F3', 'F26': 'F20'}

{'C1': 'C2', 'C2': 'C1'}

Less

{'C2': 'Less', 'C1': 'More'}

LogisticRegression

C2

Music Concert Attendance

C2 is the label picked by the algorithm with about 82.06% certainty, since the prediction likelihood of C1 is only 17.94%. F13, F11, F20, and F15 all contribute significantly to the above classification output and among them, the features that support the most positive contribution to the C2 prediction are F15, F13, and F11, while F20 drives the final prediction against assigning C2 in support of C1. F9 also contributes positively to the classification here, but F16 contributes negatively and like F20 favours C1. Finally, according to the analysis, F17, F2, F14, and F3 all have little effect on the final prediction made by the algorithm for this case.

{'C1': '17.94%', 'C2': '82.06%'}

{'F15': 'X11', 'F13': 'X1', 'F20': 'X13', 'F11': 'X3', 'F16': 'X8', 'F9': 'X6', 'F18': 'X2', 'F19': 'X9', 'F5': 'X17', 'F12': 'X10', 'F6': 'X4', 'F7': 'X14', 'F4': 'X20', 'F1': 'X18', 'F10': 'X19', 'F8': 'X7', 'F14': 'X12', 'F17': 'X15', 'F3': 'X16', 'F2': 'X5'}

{'F11': 'F15', 'F1': 'F13', 'F13': 'F20', 'F3': 'F11', 'F8': 'F16', 'F6': 'F9', 'F2': 'F18', 'F9': 'F19', 'F17': 'F5', 'F10': 'F12', 'F4': 'F6', 'F14': 'F7', 'F20': 'F4', 'F18': 'F1', 'F19': 'F10', 'F7': 'F8', 'F12': 'F14', 'F15': 'F17', 'F16': 'F3', 'F5': 'F2'}

{'C1': 'C1', 'C2': 'C2'}

> 10k

{'C1': '< 10k', 'C2': '> 10k'}

LogisticRegression

C1

Flight Price-Range Classification

The model is confident in its prediction, as it predicted class C1 with a likelihood of 90.48% and hence, for the given case, there is a smaller chance of it being any other class label. F2 and F3 are deemed the most important features whereas on the other hand all the other features have moderate to minimal amounts of influence. Both F2 and F3 have the same direction of impact, increasing the odds of the predicted label, C1. While F7 and F9 are both encouraging the model to make a prediction of C1, the others F10, F6, and F12 is pushing the model towards a different label. Many features have moderately low impact on the final prediction, but the features F4, F12, and F8 are those with the smallest influence.

{'C1': '90.48%', 'C3': '9.51%', 'C2': '0.01%'}

{'F2': 'Total_Stops', 'F3': 'Airline', 'F7': 'Destination', 'F9': 'Arrival_hour', 'F10': 'Source', 'F11': 'Duration_hours', 'F1': 'Dep_hour', 'F6': 'Dep_minute', 'F5': 'Arrival_minute', 'F4': 'Journey_month', 'F12': 'Journey_day', 'F8': 'Duration_mins'}

{'F12': 'F2', 'F9': 'F3', 'F11': 'F7', 'F5': 'F9', 'F10': 'F10', 'F7': 'F11', 'F3': 'F1', 'F4': 'F6', 'F6': 'F5', 'F2': 'F4', 'F1': 'F12', 'F8': 'F8'}

{'C1': 'C1', 'C3': 'C3', 'C2': 'C2'}

Low

{'C1': 'Low', 'C3': 'Moderate', 'C2': 'High'}

DecisionTreeClassifier

C1

Airline Passenger Satisfaction

Based on the probability distribution across the classes, the classifier is shown to have a moderately high confidence level in the C1 label assignment, with its likelihood equal to 65.0%, whereas that of C2 is only 35.0%. The prediction decision above is predominantly due to the influence of the variables F10, F15, F17, and F4. On the lower end are the least relevant variables, F13, F24, F5, F23, F3, and F16, with little to no influence on the classifier when assigning a label to the given instance. On the one hand, the top positive variables are F10, F15, and F17, increasing the probability that C1 is the correct label. Also, the top negative variables are F4, F14, F1, and F26, decreasing the classifier's response and consequently shifting the prediction verdict in the opposite direction towards C2. Other variables with a positive direction of influence are F9, F7, F20, F8, F6, F19, F11, F2, and F25.

{'C2': '35.00%', 'C1': '65.00%'}

{'F10': 'X8', 'F15': 'X2', 'F17': 'X1', 'F4': 'X21', 'F9': 'X25', 'F14': 'X10', 'F1': 'X3', 'F26': 'X9', 'F7': 'X15', 'F18': 'X7', 'F12': 'X20', 'F22': 'X12', 'F20': 'X24', 'F8': 'X6', 'F6': 'X17', 'F19': 'X23', 'F21': 'X11', 'F11': 'X22', 'F2': 'X4', 'F25': 'X14', 'F13': 'X19', 'F24': 'X18', 'F5': 'X16', 'F23': 'X13', 'F3': 'X5', 'F16': 'X26'}

{'F8': 'F10', 'F2': 'F15', 'F1': 'F17', 'F21': 'F4', 'F25': 'F9', 'F10': 'F14', 'F3': 'F1', 'F9': 'F26', 'F15': 'F7', 'F7': 'F18', 'F20': 'F12', 'F12': 'F22', 'F24': 'F20', 'F6': 'F8', 'F17': 'F6', 'F23': 'F19', 'F11': 'F21', 'F22': 'F11', 'F4': 'F2', 'F14': 'F25', 'F19': 'F13', 'F18': 'F24', 'F16': 'F5', 'F13': 'F23', 'F5': 'F3', 'F26': 'F16'}

{'C1': 'C2', 'C2': 'C1'}

Acceptable

{'C2': 'neutral or dissatisfied', 'C1': 'satisfied'}

KNeighborsClassifier

C1

Company Bankruptcy Prediction

The model's output labelling judgement for the case under consideration is as follows: C2 cannot be the label for the given case; C1 is the most likely class label with a 100.0% confidence level. The key driving factors resulting in the aforementioned classification are the values of the input features: F63, F74, F14, F24, F38, F70, and F42. F73, F31, F45, F71, F81, F4, F10, F77, F16, F33, F27, F41, and F21 are the features that have a modest effect on the decision. Aside from the aforementioned input features, all others, such as F62, F9, F49, and F69, are revealed to be irrelevant to the conclusion reached here. Not all of the influential features support labelling the current instance as C1, and they are referred to as negative features. F42, F31, F27, F41, and F21 are the negative attributes that diminish the likelihood that C1 is the correct label in this case. F63, F74, F14, and F24 are important positive features that strongly increase the likelihood that C1 is the correct label.

{'C1': '100.00%', 'C2': '0.00%'}

{'F63': ' Interest Coverage Ratio (Interest expense to EBIT)', 'F74': ' Net Income to Total Assets', 'F14': ' Realized Sales Gross Profit Growth Rate', 'F24': ' Accounts Receivable Turnover', 'F38': ' Operating Expense Rate', 'F42': ' Contingent liabilities\\/Net worth', 'F70': ' Non-industry income and expenditure\\/revenue', 'F73': ' Current Ratio', 'F31': ' Cash Flow to Liability', 'F45': ' Fixed Assets Turnover Frequency', 'F81': ' Regular Net Profit Growth Rate', 'F71': ' Quick Asset Turnover Rate', 'F4': ' Net Value Per Share (C)', 'F10': ' Operating Profit Growth Rate', 'F77': ' After-tax Net Profit Growth Rate', 'F16': ' Continuous Net Profit Growth Rate', 'F33': ' Net Value Per Share (B)', 'F27': ' Equity to Long-term Liability', 'F41': ' CFO to Assets', 'F21': ' Total debt\\/Total net worth', 'F62': ' Current Asset Turnover Rate', 'F9': " Net Income to Stockholder's Equity", 'F49': ' Operating Gross Margin', 'F69': ' Operating Profit Per Share (Yuan ¥)', 'F65': ' Operating Profit Rate', 'F36': ' Cash Flow Per Share', 'F90': ' Total income\\/Total expense', 'F68': ' No-credit Interval', 'F53': ' Liability to Equity', 'F2': ' Working Capital to Total Assets', 'F92': ' Working Capital\\/Equity', 'F12': ' Long-term Liability to Current Assets', 'F35': ' Interest-bearing debt interest rate', 'F91': ' Inventory and accounts receivable\\/Net value', 'F34': ' Realized Sales Gross Margin', 'F6': ' Current Liability to Equity', 'F32': ' Equity to Liability', 'F86': ' Current Liability to Liability', 'F75': ' Operating profit\\/Paid-in capital', 'F44': ' Operating Funds to Liability', 'F83': ' Current Liability to Current Assets', 'F85': ' Net worth\\/Assets', 'F19': ' Tax rate (A)', 'F64': ' Quick Assets\\/Current Liability', 'F66': ' After-tax net Interest Rate', 'F40': ' Per Share Net profit before tax (Yuan ¥)', 'F39': ' Total Asset Turnover', 'F48': ' Cash Reinvestment %', 'F11': ' Fixed Assets to Assets', 'F17': ' Working capitcal Turnover Rate', 'F28': ' Net profit before tax\\/Paid-in capital', 'F29': ' Net Worth Turnover Rate (times)', 'F46': ' Debt ratio %', 'F87': ' Cash Flow to Equity', 'F13': ' Long-term fund suitability ratio (A)', 'F26': ' Cash Flow to Sales', 'F55': ' Total Asset Growth Rate', 'F61': ' Inventory\\/Current Liability', 'F37': ' Allocation rate per person', 'F82': ' Inventory Turnover Rate (times)', 'F93': ' Operating profit per person', 'F60': ' Net Value Growth Rate', 'F52': ' Interest Expense Ratio', 'F50': ' ROA(B) before interest and depreciation after tax', 'F88': ' Continuous interest rate (after tax)', 'F43': ' Inventory\\/Working Capital', 'F3': ' Retained Earnings to Total Assets', 'F78': ' Total assets to GNP price', 'F23': ' Persistent EPS in the Last Four Seasons', 'F1': ' Quick Ratio', 'F79': ' Revenue per person', 'F15': ' Borrowing dependency', 'F58': ' Cash\\/Total Assets', 'F47': ' ROA(A) before interest and % after tax', 'F56': ' ROA(C) before interest and depreciation before interest', 'F30': ' Average Collection Days', 'F89': ' Current Liabilities\\/Liability', 'F7': ' Cash Flow to Total Assets', 'F22': ' Pre-tax net Interest Rate', 'F67': ' Current Liability to Assets', 'F8': ' Quick Assets\\/Total Assets', 'F20': ' Total expense\\/Assets', 'F84': ' Net Value Per Share (A)', 'F57': ' Current Assets\\/Total Assets', 'F59': ' Research and development expense rate', 'F80': ' Current Liabilities\\/Equity', 'F5': ' Cash flow rate', 'F51': ' Total Asset Return Growth Rate Ratio', 'F54': ' Degree of Financial Leverage (DFL)', 'F18': ' Cash Turnover Rate', 'F72': ' Cash\\/Current Liability', 'F25': ' Revenue Per Share (Yuan ¥)', 'F76': ' Gross Profit to Sales'}

{'F60': 'F63', 'F16': 'F74', 'F38': 'F14', 'F2': 'F24', 'F19': 'F38', 'F64': 'F42', 'F4': 'F70', 'F82': 'F73', 'F50': 'F31', 'F22': 'F45', 'F85': 'F81', 'F33': 'F71', 'F88': 'F4', 'F43': 'F10', 'F80': 'F77', 'F54': 'F16', 'F27': 'F33', 'F23': 'F27', 'F76': 'F41', 'F7': 'F21', 'F61': 'F62', 'F59': 'F9', 'F62': 'F49', 'F63': 'F69', 'F58': 'F65', 'F65': 'F36', 'F57': 'F90', 'F56': 'F68', 'F66': 'F53', 'F67': 'F2', 'F68': 'F92', 'F69': 'F12', 'F1': 'F35', 'F70': 'F91', 'F83': 'F34', 'F92': 'F6', 'F91': 'F32', 'F90': 'F86', 'F89': 'F75', 'F87': 'F44', 'F86': 'F83', 'F84': 'F85', 'F81': 'F19', 'F71': 'F64', 'F79': 'F66', 'F78': 'F40', 'F77': 'F39', 'F75': 'F48', 'F74': 'F11', 'F73': 'F17', 'F72': 'F28', 'F55': 'F29', 'F47': 'F46', 'F53': 'F87', 'F52': 'F13', 'F25': 'F26', 'F24': 'F55', 'F21': 'F61', 'F20': 'F37', 'F18': 'F82', 'F17': 'F93', 'F15': 'F60', 'F14': 'F52', 'F13': 'F50', 'F12': 'F88', 'F11': 'F43', 'F10': 'F3', 'F9': 'F78', 'F8': 'F23', 'F6': 'F1', 'F5': 'F79', 'F3': 'F15', 'F26': 'F58', 'F28': 'F47', 'F29': 'F56', 'F41': 'F30', 'F51': 'F89', 'F49': 'F7', 'F48': 'F22', 'F46': 'F67', 'F45': 'F8', 'F44': 'F20', 'F42': 'F84', 'F40': 'F57', 'F30': 'F59', 'F39': 'F80', 'F37': 'F5', 'F36': 'F51', 'F35': 'F54', 'F34': 'F18', 'F32': 'F72', 'F31': 'F25', 'F93': 'F76'}

{'C1': 'C1', 'C2': 'C2'}

No

{'C1': 'No', 'C2': 'Yes'}

BernoulliNB

C2

Student Job Placement

For the case under consideration, the model assigned C2 with very high confidence, since the likelihood of C1 being the right label is only 0.52% which is very small. F11, F5, F3, and F6 have a large positive impact on the model's output prediction. F3 and F6 have a moderately positive impact on the prediction of C2, while F9 has a similar impact but in the opposite direction. F4, F2, and F1 have a very low impact on classification. F10, F8, F7, and F12 have a larger but still insignificant effect. Examining the attributions indicates that there are only two features, F9 and F2, with values that contradict the prediction made here but, their impact on the model is smaller when compared to positive features such as F5, F3, and F11, which explains why the confidence level associated with this classification is high.

{'C1': '0.52%', 'C2': '99.48%'}

{'F11': 'workex', 'F5': 'specialisation', 'F3': 'ssc_p', 'F6': 'hsc_p', 'F9': 'degree_p', 'F10': 'gender', 'F12': 'degree_t', 'F7': 'etest_p', 'F8': 'hsc_b', 'F2': 'hsc_s', 'F1': 'ssc_b', 'F4': 'mba_p'}

{'F11': 'F11', 'F12': 'F5', 'F1': 'F3', 'F2': 'F6', 'F3': 'F9', 'F6': 'F10', 'F10': 'F12', 'F4': 'F7', 'F8': 'F8', 'F9': 'F2', 'F7': 'F1', 'F5': 'F4'}

{'C2': 'C1', 'C1': 'C2'}

Placed

{'C1': 'Not Placed', 'C2': 'Placed'}

LogisticRegression

C1

Flight Price-Range Classification

Since the likelihood of C1 being the true label is shown by the prediction algorithm outputs to be equal to 93.02 percent, there is only a small chance that the true label for the given data instance is any of the other class labels, C3 and C2. The features F9, F1, F6, and F10 are the most important ones driving the label assignment verdict above, and on the other hand, the least relevant features are shown to be F7, F12, and F4. Considering the direction of influence of each input feature, as shown by the attribution analysis, it can be concluded that the positive features steering the prediction higher towards C1 are F9, F1, F10, F6, F2, F3, and F12. The marginal doubt in the predicted output decision is attributed to the negative contributions of F11, F8, F4, F7, and F5. Considering the attributions of the features and predicted probabilities across the classes, it can be concluded that the joint positive contribution outranks the negative contributions; hence, the algorithm is confident that C1 is likely the true label.

{'C1': '93.02%', 'C3': '6.97%', 'C2': '0.01%'}

{'F9': 'Total_Stops', 'F1': 'Airline', 'F10': 'Destination', 'F6': 'Journey_day', 'F11': 'Source', 'F8': 'Dep_hour', 'F2': 'Duration_hours', 'F5': 'Dep_minute', 'F3': 'Duration_mins', 'F7': 'Arrival_minute', 'F12': 'Arrival_hour', 'F4': 'Journey_month'}

{'F12': 'F9', 'F9': 'F1', 'F11': 'F10', 'F1': 'F6', 'F10': 'F11', 'F3': 'F8', 'F7': 'F2', 'F4': 'F5', 'F8': 'F3', 'F6': 'F7', 'F5': 'F12', 'F2': 'F4'}

{'C3': 'C1', 'C2': 'C3', 'C1': 'C2'}

Low

{'C1': 'Low', 'C3': 'Moderate', 'C2': 'High'}

RandomForestClassifier

C2

Cab Surge Pricing System

Between the three possible classes, there is an 88.0% probability that the correct label for this case is C2. This means that there is a 12.0% chance that the label could be one of the other possible labels, C3 or C1. Increasing the odds of the predicted label are the variables F3, F9, F12, and F1. The next set of variables, F8, F6, and F10, have values that moderately decrease the likelihood of C2 being the correct label. F2, F11, and F7 are the other negatively contributing features, and given that they are lowly ranked, they have a marginal impact when determining the correct label for this case. The other positive features further increasing the probability that C2 is the right label are F4 and F5. Overall, we can conclude that the decision to label the case as C2 is largely due to the strong positive influence of F9, F3, F1, and F12.

{'C3': '3.00%', 'C1': '9.00%', 'C2': '88.00%'}

{'F3': 'Type_of_Cab', 'F9': 'Destination_Type', 'F12': 'Cancellation_Last_1Month', 'F1': 'Trip_Distance', 'F8': 'Customer_Rating', 'F6': 'Life_Style_Index', 'F10': 'Var3', 'F4': 'Var1', 'F5': 'Customer_Since_Months', 'F2': 'Var2', 'F11': 'Gender', 'F7': 'Confidence_Life_Style_Index'}

{'F2': 'F3', 'F6': 'F9', 'F8': 'F12', 'F1': 'F1', 'F7': 'F8', 'F4': 'F6', 'F11': 'F10', 'F9': 'F4', 'F3': 'F5', 'F10': 'F2', 'F12': 'F11', 'F5': 'F7'}

{'C2': 'C3', 'C3': 'C1', 'C1': 'C2'}

C3

{'C3': 'Low', 'C1': 'Medium', 'C2': 'High'}

RandomForestClassifier

C1

Wine Quality Prediction

Based on the input variables, the model is moderately confident that the C1 is the appropriate label for the data under consideration. As a matter of fact, the prediction likelihood associated with class C2 is about 30.42%. The preceeding classification verdict can be largely blamed on the contributions of variables F2, F8, F11, and F1, whereas those with marginally lower contributions are F4, F9, and F6. The variables with moderate contributions are F5, F7, F3, and F10. Considering their respective contributions, F2, F11, F1, and F10 are the variables with positive influence that increase the chances of C1 being the correct label for the given data. The little doubt in the label choice here could be attributed to the negative variables, mainly F8, F5, F3, and F7, which decrease the chances of the model labelling the data given as C1 since these negative variables favour selecting the alternative label, C2 over C1. Given that majority of top variables contribute positively, it is not unexpected that C1 is the picked label with reasonably high confidence.

{'C2': '30.42%', 'C1': '69.58%'}

{'F2': 'alcohol', 'F8': 'sulphates', 'F11': 'volatile acidity', 'F1': 'total sulfur dioxide', 'F5': 'fixed acidity', 'F7': 'citric acid', 'F3': 'residual sugar', 'F10': 'density', 'F4': 'chlorides', 'F9': 'pH', 'F6': 'free sulfur dioxide'}

{'F11': 'F2', 'F10': 'F8', 'F2': 'F11', 'F7': 'F1', 'F1': 'F5', 'F3': 'F7', 'F4': 'F3', 'F8': 'F10', 'F5': 'F4', 'F9': 'F9', 'F6': 'F6'}

{'C1': 'C2', 'C2': 'C1'}

high quality

{'C2': 'low_quality', 'C1': 'high quality'}

SVC

C2

E-Commerce Shipping

The classifier is 69.02% certain that the given case is under the class label C2, implying that the likelihood of C1 is only 30.98%. Analysis performed to understand the contribution of each input feature revealed that: F6, F4, and F2 are the most influential features when assigning a label to the given case. Features F5, F3, F7, and F8 have moderate contributions, whereas the F10, F1 and F9 have lower relevance to the final classification decision. F6 and F2 push the class assignment towards C2, whereas F4 does the opposite, decreasing the likelihood of C2. Similar to F4, F5, and F3 negatively impact the C2 classification, whereas F8, F10, and F7 positively push the decision towards the C2 class. Features F1, and F9 all have little impact on the final decision, with F9 having the least impact.

{'C2': '69.02%', 'C1': '30.98%'}

{'F6': 'Weight_in_gms', 'F4': 'Discount_offered', 'F2': 'Prior_purchases', 'F5': 'Customer_care_calls', 'F3': 'Cost_of_the_Product', 'F7': 'Mode_of_Shipment', 'F8': 'Customer_rating', 'F10': 'Gender', 'F1': 'Product_importance', 'F9': 'Warehouse_block'}

{'F3': 'F6', 'F2': 'F4', 'F8': 'F2', 'F6': 'F5', 'F1': 'F3', 'F5': 'F7', 'F7': 'F8', 'F10': 'F10', 'F9': 'F1', 'F4': 'F9'}

{'C1': 'C2', 'C2': 'C1'}

On-time

{'C2': 'On-time', 'C1': 'Late'}

RandomForestClassifier

C2

Advertisement Prediction

The classifier trained on this prediction problem assigns a label to a given case based on the information supplied. The class assigned by the classifier to the case under consideration is C2. The probability that C1 is the correct label is around 25.28%; therefore, it is less likely to be the true label. The above classification decision is mainly based on the influence of the features F3, F6, F1, F4, F2, F7, and F5. Of the above stated features, F4 and F6 are the ones shown to have a negative impact, decreasing the odds of C2 being the accurate label for the given case and encouraging the classifier to select C1 instead. Finally, it can be concluded that there is a moderately high level of confidence in the assigned label, which can be attributed to the strong positive contribution of F3 combined with other positive features such as F1 and F2.

{'C2': '74.72%', 'C1': '25.28%'}

{'F3': 'Daily Time Spent on Site', 'F6': 'Daily Internet Usage', 'F1': 'Age', 'F4': 'ad_day', 'F2': 'Area Income', 'F7': 'Gender', 'F5': 'ad_month'}

{'F1': 'F3', 'F4': 'F6', 'F2': 'F1', 'F7': 'F4', 'F3': 'F2', 'F5': 'F7', 'F6': 'F5'}

{'C1': 'C2', 'C2': 'C1'}

Skip

{'C2': 'Skip', 'C1': 'Watch'}

GradientBoostingClassifier

C1

Food Ordering Customer Churn Prediction

The case given is labelled as C1 by the classifier with a confidence level equal to 82.07%. Therefore, the probability of C2 being the correct label is only 17.93%. The classification above is mainly due to the contributions of features such as F29, F3, F40, and F8. F45, F25, and F34 are the next three with moderate influence. However, not all the features are considered by the classifier when determining the correct label for the given case. F2, F44, F38, and F27 are notable irrelevant features. With regards to the direction of influence of the relevant features, F29, F3, F40, and F8 are the top features with strong positive contributions favouring the assignment of label C1. The top negative features that shift the classification in a different direction are F45, F25, F39, and F9. Considering the fact that a number of the relevant features have positive attributions, it is not surprising that the classifier is quite certain that the appropriate label is C1 instead of C2.

{'C2': '17.93%', 'C1': '82.07%'}

{'F29': 'More restaurant choices', 'F3': 'Ease and convenient', 'F40': 'Bad past experience', 'F8': 'Time saving', 'F45': 'Easy Payment option', 'F25': 'Good Tracking system', 'F34': 'Wrong order delivered', 'F39': 'Influence of rating', 'F41': 'Late Delivery', 'F9': 'Less Delivery time', 'F11': 'Long delivery time', 'F15': 'Delivery person ability', 'F19': 'Order placed by mistake', 'F26': 'More Offers and Discount', 'F35': 'Freshness ', 'F23': 'Unavailability', 'F20': 'Delay of delivery person picking up food', 'F24': 'Poor Hygiene', 'F21': 'Order Time', 'F43': 'Delay of delivery person getting assigned', 'F2': 'High Quality of package', 'F44': 'Residence in busy location', 'F38': 'Good Taste ', 'F27': 'Temperature', 'F4': 'Google Maps Accuracy', 'F1': 'Good Road Condition', 'F36': 'Number of calls', 'F12': 'Low quantity low time', 'F37': 'Politeness', 'F28': 'Maximum wait time', 'F7': 'Age', 'F42': 'Influence of time', 'F13': 'Missing item', 'F30': 'Family size', 'F16': 'Unaffordable', 'F5': 'Health Concern', 'F17': 'Self Cooking', 'F14': 'Good Food quality', 'F18': 'Perference(P2)', 'F10': 'Perference(P1)', 'F32': 'Educational Qualifications', 'F6': 'Monthly Income', 'F46': 'Occupation', 'F33': 'Marital Status', 'F31': 'Gender', 'F22': 'Good Quantity'}

{'F12': 'F29', 'F10': 'F3', 'F21': 'F40', 'F11': 'F8', 'F13': 'F45', 'F16': 'F25', 'F27': 'F34', 'F38': 'F39', 'F19': 'F41', 'F39': 'F9', 'F24': 'F11', 'F37': 'F15', 'F29': 'F19', 'F14': 'F26', 'F43': 'F35', 'F22': 'F23', 'F26': 'F20', 'F20': 'F24', 'F31': 'F21', 'F25': 'F43', 'F40': 'F2', 'F33': 'F44', 'F45': 'F38', 'F44': 'F27', 'F34': 'F4', 'F35': 'F1', 'F41': 'F36', 'F36': 'F12', 'F42': 'F37', 'F32': 'F28', 'F1': 'F7', 'F30': 'F42', 'F28': 'F13', 'F7': 'F30', 'F23': 'F16', 'F18': 'F5', 'F17': 'F17', 'F15': 'F14', 'F9': 'F18', 'F8': 'F10', 'F6': 'F32', 'F5': 'F6', 'F4': 'F46', 'F3': 'F33', 'F2': 'F31', 'F46': 'F22'}

{'C2': 'C2', 'C1': 'C1'}

Go Away

{'C2': 'Return', 'C1': 'Go Away'}

SVM_linear

C4

Mobile Price-Range Classification

According to the algorithm, there is little to no chance that the correct label for the given data instance is any of the following classes: C3, C2, and C1. It is very confident that the proper label is C4. This label assignment is largely due to the parts played by the features F8, F1, and F9. On the lower end are the input features F16, F14, F15, and F7, which are shown to be less relevant when it comes to this labelling assignment task. Finally, among the top features identified during the attribution investogation, only F6 and F20 are features with a negative influence, decreasing the odds of C4 being the appropriate label here.

{'C3': '0.00%', 'C2': '0.00%', 'C1': '0.00%', 'C4': '100.00%'}

{'F8': 'ram', 'F1': 'battery_power', 'F9': 'px_width', 'F20': 'int_memory', 'F6': 'sc_h', 'F11': 'pc', 'F10': 'mobile_wt', 'F19': 'fc', 'F18': 'n_cores', 'F3': 'clock_speed', 'F12': 'blue', 'F4': 'three_g', 'F17': 'touch_screen', 'F2': 'm_dep', 'F13': 'px_height', 'F5': 'talk_time', 'F16': 'dual_sim', 'F14': 'wifi', 'F15': 'four_g', 'F7': 'sc_w'}

{'F11': 'F8', 'F1': 'F1', 'F10': 'F9', 'F4': 'F20', 'F12': 'F6', 'F8': 'F11', 'F6': 'F10', 'F3': 'F19', 'F7': 'F18', 'F2': 'F3', 'F15': 'F12', 'F18': 'F4', 'F19': 'F17', 'F5': 'F2', 'F9': 'F13', 'F14': 'F5', 'F16': 'F16', 'F20': 'F14', 'F17': 'F15', 'F13': 'F7'}

{'C4': 'C3', 'C1': 'C2', 'C3': 'C1', 'C2': 'C4'}

r4

{'C3': 'r1', 'C2': 'r2', 'C1': 'r3', 'C4': 'r4'}

SVC

C2

Paris House Classification

The model predicts that the label for this case is C2 with a high degree of certainty of about 99.19% and the probability of the other label is only 0.81%. From the analysis, the variables with the strongest attributions to this classification decision are F14, F12, and F4. The attributions of these variables increased the response of the model in favour of labelling the case as C2. Other variables that positively supported the label decision include F8, F7, and F3. Not all the variables support the model's prediction of C2 and this is because the values of F16, F17, F13, F10, and F6 are driving the prediction towards C1. The joint attribution from these variables is weaker than that from F14, F12, and F4, so the model is biased toward predicting C2. Finally, F1, F2, F9, and F5 are the least important positive features, given that they have minimal attributions in favour of C2.

{'C2': '99.19%', 'C1': '0.81%'}

{'F14': 'isNewBuilt', 'F4': 'hasYard', 'F12': 'hasPool', 'F16': 'hasStormProtector', 'F17': 'hasStorageRoom', 'F8': 'made', 'F7': 'basement', 'F3': 'numberOfRooms', 'F13': 'squareMeters', 'F11': 'floors', 'F10': 'numPrevOwners', 'F15': 'garage', 'F6': 'attic', 'F1': 'cityCode', 'F2': 'price', 'F9': 'cityPartRange', 'F5': 'hasGuestRoom'}

{'F3': 'F14', 'F1': 'F4', 'F2': 'F12', 'F4': 'F16', 'F5': 'F17', 'F12': 'F8', 'F13': 'F7', 'F7': 'F3', 'F6': 'F13', 'F8': 'F11', 'F11': 'F10', 'F15': 'F15', 'F14': 'F6', 'F9': 'F1', 'F17': 'F2', 'F10': 'F9', 'F16': 'F5'}

{'C2': 'C2', 'C1': 'C1'}

Basic

{'C2': 'Basic', 'C1': 'Luxury'}

LogisticRegression

C1

Used Cars Price-Range Prediction

According to the output prediction probabilities across the two classes, the output decision for the given data is C1 with a very high confidence level. C2 has a prediction probability of about 0.00%. The variables contributing most to the abovementioned classification are F2, F10, and F1, whereas F3 and F4 are the least influential variables. The very high confidence level associated with the classification decision here could be attributed to the fact that a greater number of the input variables have attributions that increase the model's response towards label C1. F7, F9, and F3 are the variables with negative contributions that attempt to push the model to label this case as C2. To put it in a nutshell, the joint contribution of the negative variables is very low unlike that of the positive variables, hence the model's certainty in the decision here.

{'C1': '100.00%', 'C2': '0.00%'}

{'F2': 'car_age', 'F10': 'Power', 'F1': 'Fuel_Type', 'F8': 'Engine', 'F5': 'Seats', 'F6': 'Transmission', 'F7': 'Kilometers_Driven', 'F9': 'Name', 'F3': 'Mileage', 'F4': 'Owner_Type'}

{'F5': 'F2', 'F4': 'F10', 'F7': 'F1', 'F3': 'F8', 'F10': 'F5', 'F8': 'F6', 'F1': 'F7', 'F6': 'F9', 'F2': 'F3', 'F9': 'F4'}

{'C2': 'C1', 'C1': 'C2'}

Low

{'C1': 'Low', 'C2': 'High'}

SVC

C1

Tic-Tac-Toe Strategy

With a labelling confidence level of 99.50%, the classifier predicts the label C1 in this situation. Hence, it is correct to conclude that the classifier is less certain that C2 is the proper label for the case here. The analysis indicates that five features contradict the decision above, while four features support the classifier. The features contradicting the prediction are usually referred to as negative features while those supporting it are referred to as positive features. The negative features decreasing the odds of C1 being the correct label are F4, F9, F8, F7, and F6. Conversely, the positive features increasing the odds of C1 are F3, F5, F1, and F2.

{'C2': '0.50%', 'C1': '99.50%'}

{'F4': 'middle-middle-square', 'F3': 'top-left-square', 'F5': 'bottom-left-square', 'F1': 'bottom-right-square', 'F9': ' top-right-square', 'F8': 'middle-right-square', 'F7': 'top-middle-square', 'F2': 'middle-left-square', 'F6': 'bottom-middle-square'}

{'F5': 'F4', 'F1': 'F3', 'F7': 'F5', 'F9': 'F1', 'F3': 'F9', 'F6': 'F8', 'F2': 'F7', 'F4': 'F2', 'F8': 'F6'}

{'C1': 'C2', 'C2': 'C1'}

player B win

{'C2': 'player B lose', 'C1': 'player B win'}

SVMClassifier_poly

C2

Employee Attrition

The class assigned by the model is C2 with a close to 97.67% confidence level, implying that the likelihood of C1 is only 2.33%. Based on the analysis, the most important features considered during the classification are F6, F1, F29, and F9 but among these features, F1 and F29 are the only ones with negative attributions, decreasing the likelihood of C2 being the label for the given case. Furthermore, moderately influencing the decision are F11, F10, F23, and F27. F11, F10, and F23 have positive attributions, while F27 has a negative impact, shifting the prediction in a different direction. Finally, the features with insignificant impact on the model when it comes to this case include F20, F12, F15, and F28.

{'C2': '97.67%', 'C1': '2.33%'}

{'F6': 'OverTime', 'F1': 'JobSatisfaction', 'F29': 'BusinessTravel', 'F9': 'MaritalStatus', 'F11': 'EnvironmentSatisfaction', 'F10': 'Department', 'F23': 'Age', 'F27': 'YearsInCurrentRole', 'F26': 'TotalWorkingYears', 'F3': 'WorkLifeBalance', 'F17': 'JobLevel', 'F7': 'JobInvolvement', 'F25': 'EducationField', 'F2': 'JobRole', 'F21': 'MonthlyIncome', 'F13': 'PerformanceRating', 'F4': 'DistanceFromHome', 'F14': 'Education', 'F8': 'Gender', 'F19': 'YearsWithCurrManager', 'F20': 'PercentSalaryHike', 'F12': 'RelationshipSatisfaction', 'F15': 'MonthlyRate', 'F28': 'DailyRate', 'F5': 'YearsSinceLastPromotion', 'F24': 'HourlyRate', 'F30': 'YearsAtCompany', 'F18': 'TrainingTimesLastYear', 'F16': 'StockOptionLevel', 'F22': 'NumCompaniesWorked'}

{'F26': 'F6', 'F30': 'F1', 'F17': 'F29', 'F25': 'F9', 'F28': 'F11', 'F21': 'F10', 'F1': 'F23', 'F14': 'F27', 'F11': 'F26', 'F20': 'F3', 'F5': 'F17', 'F29': 'F7', 'F22': 'F25', 'F24': 'F2', 'F6': 'F21', 'F19': 'F13', 'F3': 'F4', 'F27': 'F14', 'F23': 'F8', 'F16': 'F19', 'F9': 'F20', 'F18': 'F12', 'F7': 'F15', 'F2': 'F28', 'F15': 'F5', 'F4': 'F24', 'F13': 'F30', 'F12': 'F18', 'F10': 'F16', 'F8': 'F22'}

{'C2': 'C2', 'C1': 'C1'}

Stay

{'C2': 'Leave', 'C1': 'Leave'}

KNeighborsClassifier

C1

Advertisement Prediction

With a higher degree of confidence, the model labels this given case as C1 since there is a zero chance that it is C2. The classification here can be attributed to all the features having positive contributions, decreasing the odds of C2 being the correct label. The features can be ranked based on their degree of influence from the most relevant to the least relevant as follows: F5, F6, F3, F1, F7, F4, F2. This implies that F5 is the most influential feature, while F2 is the least influential among the input features.

{'C2': '0.00%', 'C1': '100.00%'}

{'F5': 'Daily Time Spent on Site', 'F6': 'Area Income', 'F3': 'Age', 'F1': 'Daily Internet Usage', 'F7': 'ad_day', 'F4': 'Gender', 'F2': 'ad_month'}

{'F1': 'F5', 'F3': 'F6', 'F2': 'F3', 'F4': 'F1', 'F7': 'F7', 'F5': 'F4', 'F6': 'F2'}

{'C2': 'C2', 'C1': 'C1'}

Watch

{'C2': 'Skip', 'C1': 'Watch'}

SVM_poly

C2

Mobile Price-Range Classification

According to the model, C2 has a prediction probability of 99.45 percent, C3 has a prediction probability of 0.47 percent, C1 has a prediction probability of 0.04 percent, and C4 has a prediction probability of 0.05 percent, therefore, the most likely class is C2. F13 and F19 positively influence the above-mentioned label decision in favour of C2, but F9 has the opposite effect, favouring a different label. F16 and F8 both have a similar negative impact on the C2 prediction, whereas F20 has a positive impact. In this case, F11, F7, F17, and F5 have little influence on the labelling result. All in all, the model is confident in its assignment of the C2 class as shown by the predicted probabilities across the classes.

{'C2': '99.45%', 'C3': '0.47%', 'C1': '0.04%', 'C4': '0.05%'}

{'F19': 'ram', 'F13': 'battery_power', 'F9': 'px_height', 'F16': 'px_width', 'F20': 'dual_sim', 'F8': 'four_g', 'F11': 'touch_screen', 'F7': 'int_memory', 'F17': 'pc', 'F5': 'n_cores', 'F2': 'fc', 'F15': 'clock_speed', 'F18': 'three_g', 'F4': 'sc_w', 'F6': 'wifi', 'F3': 'm_dep', 'F14': 'mobile_wt', 'F10': 'talk_time', 'F1': 'sc_h', 'F12': 'blue'}

{'F11': 'F19', 'F1': 'F13', 'F9': 'F9', 'F10': 'F16', 'F16': 'F20', 'F17': 'F8', 'F19': 'F11', 'F4': 'F7', 'F8': 'F17', 'F7': 'F5', 'F3': 'F2', 'F2': 'F15', 'F18': 'F18', 'F13': 'F4', 'F20': 'F6', 'F5': 'F3', 'F6': 'F14', 'F14': 'F10', 'F12': 'F1', 'F15': 'F12'}

{'C1': 'C2', 'C2': 'C3', 'C4': 'C1', 'C3': 'C4'}

r1

{'C2': 'r1', 'C3': 'r2', 'C1': 'r3', 'C4': 'r4'}

GradientBoostingClassifier

C1

Food Ordering Customer Churn Prediction

Per the model employed here, the prediction probability of C2 is only 17.93%, and that of C1 is equal to 82.07%. Given the information provided to the model, the most valid conclusion regarding the true label is that C1 is without a doubt the most likely one. The attributions analysis indicates that F36, F32, F3, F30, and F14 are the major drivers resulting in the prediction probabilities across the classes under consideration. At the tail end are features such as F28, F8, F41, and F31 that have very little influence on the decision made with respect to the given case. Among the influential features, only F36, F32, F30, F6, F1, F2, F27, F35, and F46 have positive contributions in support of labelling the given case as C1. On the other hand, the negative features such as F3, F14, F15, F17, F4, F25, and F34, suggest C2 could likely be the true label in this case. Overall, the marginal doubt in the correctness of assigning C1 to the case under consideration is attributed to the negative features driving the model's decision in the direction of C2 away from C1. But the higher influence of positive features such as F36 and F32 ensures that C1 is assigned as the most probable label.

{'C1': '82.07%', 'C2': '17.93%'}

{'F36': 'Ease and convenient', 'F32': 'More restaurant choices', 'F3': 'Bad past experience', 'F30': 'More Offers and Discount', 'F14': 'Unavailability', 'F15': 'Good Food quality', 'F6': 'Low quantity low time', 'F1': 'Delay of delivery person getting assigned', 'F17': 'Late Delivery', 'F2': 'Less Delivery time', 'F34': 'Residence in busy location', 'F4': 'Freshness ', 'F25': 'Educational Qualifications', 'F27': 'Influence of rating', 'F38': 'Occupation', 'F10': 'Perference(P1)', 'F12': 'Delivery person ability', 'F35': 'Good Taste ', 'F46': 'Long delivery time', 'F5': 'Self Cooking', 'F28': 'Influence of time', 'F8': 'High Quality of package', 'F41': 'Number of calls', 'F31': 'Good Road Condition', 'F42': 'Politeness', 'F43': 'Google Maps Accuracy', 'F24': 'Temperature', 'F16': 'Maximum wait time', 'F7': 'Order Time', 'F22': 'Age', 'F19': 'Order placed by mistake', 'F26': 'Missing item', 'F23': 'Wrong order delivered', 'F18': 'Delay of delivery person picking up food', 'F11': 'Family size', 'F39': 'Unaffordable', 'F45': 'Poor Hygiene', 'F20': 'Health Concern', 'F33': 'Good Tracking system', 'F44': 'Easy Payment option', 'F9': 'Time saving', 'F37': 'Perference(P2)', 'F13': 'Monthly Income', 'F21': 'Marital Status', 'F29': 'Gender', 'F40': 'Good Quantity'}

{'F10': 'F36', 'F12': 'F32', 'F21': 'F3', 'F14': 'F30', 'F22': 'F14', 'F15': 'F15', 'F36': 'F6', 'F25': 'F1', 'F19': 'F17', 'F39': 'F2', 'F33': 'F34', 'F43': 'F4', 'F6': 'F25', 'F38': 'F27', 'F4': 'F38', 'F8': 'F10', 'F37': 'F12', 'F45': 'F35', 'F24': 'F46', 'F17': 'F5', 'F30': 'F28', 'F40': 'F8', 'F41': 'F41', 'F35': 'F31', 'F42': 'F42', 'F34': 'F43', 'F44': 'F24', 'F32': 'F16', 'F31': 'F7', 'F1': 'F22', 'F29': 'F19', 'F28': 'F26', 'F27': 'F23', 'F26': 'F18', 'F7': 'F11', 'F23': 'F39', 'F20': 'F45', 'F18': 'F20', 'F16': 'F33', 'F13': 'F44', 'F11': 'F9', 'F9': 'F37', 'F5': 'F13', 'F3': 'F21', 'F2': 'F29', 'F46': 'F40'}

{'C1': 'C1', 'C2': 'C2'}

Return

{'C1': 'Return', 'C2': 'Go Away'}

RandomForestClassifier

C2

Personal Loan Modelling

The model is about 90.0% certain or sure that the correct label based on the input features of the given case is C2. The features with the most significant influence on the decision are F9, F3, F1, and F7. The influence of the features can be categorised as positive or negative traits depending on the direction of the effect on the model. Positive features increase the likelihood of the most likely class (i.e., C2), whereas negative features reduce the model's responsiveness to the assigned label, favouring the less likely class (i.e., C1). From the attribution analysis, F5, F2, and F8 are the negative features here. Overall, the negative features are shown to have moderate to low influence compared to the positive features, hence explaining why the model is very confident about the assigned label C2.

{'C1': '10.00%', 'C2': '90.00%'}

{'F3': 'Income', 'F9': 'CCAvg', 'F1': 'CD Account', 'F7': 'Education', 'F2': 'Extra_service', 'F4': 'Securities Account', 'F6': 'Family', 'F5': 'Mortgage', 'F8': 'Age'}

{'F2': 'F3', 'F4': 'F9', 'F8': 'F1', 'F5': 'F7', 'F9': 'F2', 'F7': 'F4', 'F3': 'F6', 'F6': 'F5', 'F1': 'F8'}

{'C2': 'C1', 'C1': 'C2'}

Accept

{'C1': 'Reject', 'C2': 'Accept'}

LogisticRegression

C2

Tic-Tac-Toe Strategy

With an 81.01% chance of being correct, C2 is the most likely label, consequently, the C1 class's prediction probability is only 18.99%. The algorithm or classifier got the above prediction mostly due to the influence of features like F2, F8, F9, and F4. F3, which is found to have very little impact with regard to the label choice here, is the least relevant feature for the algorithm. F8, F6, F9, and F4 have a positive direction of influence, pushing the algorithm higher towards the C2 label. Negative features like F2, F1, and F5 favour choosing or labelling the case as C1.

{'C1': '18.99%', 'C2': '81.01%'}

{'F8': 'bottom-right-square', 'F2': 'middle-middle-square', 'F9': 'bottom-left-square', 'F4': 'middle-left-square', 'F6': 'top-left-square', 'F5': ' top-right-square', 'F1': 'middle-right-square', 'F7': 'top-middle-square', 'F3': 'bottom-middle-square'}

{'F9': 'F8', 'F5': 'F2', 'F7': 'F9', 'F4': 'F4', 'F1': 'F6', 'F3': 'F5', 'F6': 'F1', 'F2': 'F7', 'F8': 'F3'}

{'C2': 'C1', 'C1': 'C2'}

player B win

{'C1': 'player B lose', 'C2': 'player B win'}

SVC

C2

Student Job Placement

The model makes classification decisions based on the information provided to it and for the case here, the prediction probabilities across the two class labels, C1 and C2, are 49.32% and 50.68%, respectively. Based on these prediction probabilities, the label assigned is C2, since it has the highest likelihood, however, the model is not very certain about the correctness of the assigned label since its probability is marginally higher than the average. The uncertainty in the classification here can be blamed on the fact that only F4, F7, F3, F2, and F1 have positive attributions, shifting the decision higher towards C2. On the other hand, features F8, F9, F5, F12, F6, F10, and F11 have negative contributions that decrease the prediction likelihood of C2 while increasing that of C1. To cut a long story short, the most positive features are F4 and F7, whereas the most negative ones are F8 and F9. Finally, F6, F1, and F10 are not as important as all the previously mentioned features hence received little attention from the model.

{'C1': '49.32%', 'C2': '50.68%'}

{'F4': 'mba_p', 'F8': 'specialisation', 'F9': 'etest_p', 'F7': 'gender', 'F5': 'workex', 'F3': 'hsc_s', 'F12': 'hsc_p', 'F2': 'degree_t', 'F11': 'ssc_p', 'F6': 'degree_p', 'F1': 'ssc_b', 'F10': 'hsc_b'}

{'F5': 'F4', 'F12': 'F8', 'F4': 'F9', 'F6': 'F7', 'F11': 'F5', 'F9': 'F3', 'F2': 'F12', 'F10': 'F2', 'F1': 'F11', 'F3': 'F6', 'F7': 'F1', 'F8': 'F10'}

{'C2': 'C1', 'C1': 'C2'}

Placed

{'C1': 'Not Placed', 'C2': 'Placed'}

SVMClassifier_liner

C1

Employee Attrition

The most likely label for the given case is C1 since the predicted probability of C2 is only 34.27% and this means that the likelihood of C1 is 65.73%. The most relevant features that led to the C1 classification verdict are F11, F10, F9, F27, and F2. However, some of the features are deemed irrelevant to the above verdict and these include F14, F20, F21, and F3. Among the relevant features with some degree of impact, seven are shown to drive the model's class assignment towards the C2, while the remaining support the C1 prediction. Notable negative features swinging the prediction towards C2 are F11, F10, and F9, while the notable positive features are F27 and F2. The small uncertainty associated with the prediction decision for the given case could be attributed to the fact that all the three most important features are negative features whose values contradict assigning the label C1.

{'C1': '65.73%', 'C2': '34.27%'}

{'F11': 'OverTime', 'F10': 'NumCompaniesWorked', 'F9': 'YearsSinceLastPromotion', 'F27': 'BusinessTravel', 'F2': 'MaritalStatus', 'F12': 'RelationshipSatisfaction', 'F16': 'Department', 'F23': 'Age', 'F30': 'Gender', 'F28': 'JobInvolvement', 'F15': 'JobRole', 'F1': 'PerformanceRating', 'F19': 'EnvironmentSatisfaction', 'F7': 'DailyRate', 'F18': 'YearsAtCompany', 'F24': 'YearsWithCurrManager', 'F25': 'Education', 'F8': 'EducationField', 'F13': 'WorkLifeBalance', 'F17': 'DistanceFromHome', 'F14': 'YearsInCurrentRole', 'F20': 'TrainingTimesLastYear', 'F21': 'TotalWorkingYears', 'F3': 'StockOptionLevel', 'F6': 'PercentSalaryHike', 'F22': 'MonthlyRate', 'F29': 'MonthlyIncome', 'F5': 'JobLevel', 'F4': 'HourlyRate', 'F26': 'JobSatisfaction'}

{'F26': 'F11', 'F8': 'F10', 'F15': 'F9', 'F17': 'F27', 'F25': 'F2', 'F18': 'F12', 'F21': 'F16', 'F1': 'F23', 'F23': 'F30', 'F29': 'F28', 'F24': 'F15', 'F19': 'F1', 'F28': 'F19', 'F2': 'F7', 'F13': 'F18', 'F16': 'F24', 'F27': 'F25', 'F22': 'F8', 'F20': 'F13', 'F3': 'F17', 'F14': 'F14', 'F12': 'F20', 'F11': 'F21', 'F10': 'F3', 'F9': 'F6', 'F7': 'F22', 'F6': 'F29', 'F5': 'F5', 'F4': 'F4', 'F30': 'F26'}

{'C1': 'C1', 'C2': 'C2'}

Stay

{'C1': 'Leave', 'C2': 'Leave'}

RandomForestClassifier

C2

Printer Sales

Per the classifier for the given data, the most plausible label is C2. F4, F16, F11, and F25 are the main features pushing for the above-mentioned outcome. F3, F20, F8, F24, F15, and F6, on the other hand, have little contribution to the classifier employed here. F12, F10, F26, and F14 have a moderate contribution to the assignment of C2. The classifier's confidence in the label decision above can be attributed to larger positive attributions of F10, F12, F11, and F16 compared to the negative attributions of F26, F18, F4, F2, F25, and F9.

{'C1': '20.00%', 'C2': '80.00%'}

{'F16': 'X24', 'F11': 'X1', 'F4': 'X8', 'F25': 'X21', 'F26': 'X4', 'F10': 'X10', 'F12': 'X3', 'F14': 'X15', 'F13': 'X9', 'F5': 'X23', 'F7': 'X25', 'F22': 'X7', 'F2': 'X22', 'F21': 'X11', 'F18': 'X17', 'F17': 'X18', 'F19': 'X26', 'F9': 'X13', 'F23': 'X6', 'F1': 'X20', 'F24': 'X16', 'F6': 'X19', 'F8': 'X2', 'F3': 'X12', 'F15': 'X5', 'F20': 'X14'}

{'F24': 'F16', 'F1': 'F11', 'F8': 'F4', 'F21': 'F25', 'F4': 'F26', 'F10': 'F10', 'F3': 'F12', 'F15': 'F14', 'F9': 'F13', 'F23': 'F5', 'F25': 'F7', 'F7': 'F22', 'F22': 'F2', 'F11': 'F21', 'F17': 'F18', 'F18': 'F17', 'F26': 'F19', 'F13': 'F9', 'F6': 'F23', 'F20': 'F1', 'F16': 'F24', 'F19': 'F6', 'F2': 'F8', 'F12': 'F3', 'F5': 'F15', 'F14': 'F20'}

{'C1': 'C1', 'C2': 'C2'}

More

{'C1': 'Less', 'C2': 'More'}

SGDClassifier

C2

Job Change of Data Scientists

The least probable class, according to the classification algorithm, is C1, with a prediction probability of 25.12%, therefore, we can conclude that the algorithm is quite confident that the correct label for this data is C2. Analysing the attributions revealed that F6, F1, F10, and F11 are the most relevant features, whereas F3, F4, and F2 are the least relevant features. Increasing the algorithm's response in favour of C2 are the positive features F6, F10, F11, F4, F3, and F8. On the contrary, all the other features, F1, F12, F7, F9, F5, and F2, drive the algorithm towards labelling the given data as C1, hence they are considered negative features. Furthermore, the negative influence on the algorithm is the reason why the confidence level in the C2 is reduced to 74.88%.

{'C2': '74.88%', 'C1': '25.12%'}

{'F6': 'city_development_index', 'F10': 'relevent_experience', 'F1': 'city', 'F11': 'major_discipline', 'F12': 'experience', 'F7': 'training_hours', 'F9': 'education_level', 'F8': 'gender', 'F5': 'enrolled_university', 'F3': 'company_type', 'F4': 'last_new_job', 'F2': 'company_size'}

{'F1': 'F6', 'F5': 'F10', 'F3': 'F1', 'F8': 'F11', 'F9': 'F12', 'F2': 'F7', 'F7': 'F9', 'F4': 'F8', 'F6': 'F5', 'F11': 'F3', 'F12': 'F4', 'F10': 'F2'}

{'C1': 'C2', 'C2': 'C1'}

Stay

{'C2': 'Stay', 'C1': 'Leave'}

SVM_poly

C1

Mobile Price-Range Classification

According to the classification algorithm, neither C2 nor C3 nor C4 is the correct label for the given case. It is 100.0% certain that C1 is the right label. The higher degree of certainty in the above prediction can be attributed to the positive contributions of F20, F11, and F6. The other positive features include F4, F12, F16, and F1, however, unlike F20, F11, and F6, these features have a moderately low impact on the algorithm's decision. The remaining positive features, F17, F2, F15, and F13, are among the least influential input features considered by the algorithm. There are other features such as F7, F10, F3, and F9 whose contributions only serve to decrease the odds of C1 being the correct label for the given case. Regarding the high confidence of the algorithm with respect to this classification, one can conclude that the negative features have little influence on the algorithm's label decision here.

{'C2': '0.00%', 'C3': '0.00%', 'C4': '0.00%', 'C1': '100.00%'}

{'F20': 'ram', 'F11': 'battery_power', 'F6': 'px_width', 'F7': 'int_memory', 'F10': 'sc_h', 'F3': 'wifi', 'F4': 'fc', 'F9': 'three_g', 'F12': 'mobile_wt', 'F18': 'clock_speed', 'F19': 'm_dep', 'F5': 'n_cores', 'F16': 'pc', 'F14': 'touch_screen', 'F1': 'blue', 'F17': 'talk_time', 'F8': 'sc_w', 'F2': 'px_height', 'F15': 'four_g', 'F13': 'dual_sim'}

{'F11': 'F20', 'F1': 'F11', 'F10': 'F6', 'F4': 'F7', 'F12': 'F10', 'F20': 'F3', 'F3': 'F4', 'F18': 'F9', 'F6': 'F12', 'F2': 'F18', 'F5': 'F19', 'F7': 'F5', 'F8': 'F16', 'F19': 'F14', 'F15': 'F1', 'F14': 'F17', 'F13': 'F8', 'F9': 'F2', 'F17': 'F15', 'F16': 'F13'}

{'C1': 'C2', 'C3': 'C3', 'C2': 'C4', 'C4': 'C1'}

r4

{'C2': 'r1', 'C3': 'r2', 'C4': 'r3', 'C1': 'r4'}

DNN

C2

Ethereum Fraud Detection

The prediction likelihoods across the two classes are 15.35% for class C1 and 84.65% for C2, it can be concluded that C2 is the most probable class label for the given data instance. According to the attribution analysis conducted, the different input variables have varying degrees of influence on the model's decision here. The most influential set of variables is F9, F24, F18, F15, F28, F13, and F21, while the variables with the least influence include F19, F7, F37, F20, F22, and F26. The following or subsequent analysis performed to understand the direction of contribution of of the features will focus on the most influential ones controlling the label selection here. Among the top influential features, F9, F24, F18, F15, and F21, only F9 and F24 have negative contributions, decreasing the probability that C2 is the correct label, and they strongly support labelling the case as C1 instead. Pushing the classification decision in favour of C2 are the positive variables such as F18, F15, and F21. The contributions of the remaining variables, including F28, F13, and F36, have moderate to low influence. All in all, the marginal uncertainty in the decision here is mainly due to the negative influences of F9, F24, F10, and F5, but the positive contributions of F18, F15, F36, F28, F13, and F21 drive the decision higher towards C2.

{'C1': '15.35%', 'C2': '84.65%'}

{'F9': ' ERC20 uniq rec contract addr', 'F24': ' ERC20 uniq rec token name', 'F18': 'min value received', 'F15': 'Time Diff between first and last (Mins)', 'F21': 'avg val sent', 'F28': ' ERC20 uniq sent token name', 'F13': 'Sent tnx', 'F36': 'Avg min between received tnx', 'F10': 'Unique Received From Addresses', 'F5': ' ERC20 uniq rec addr', 'F14': 'total transactions (including tnx to create contract', 'F31': 'Avg min between sent tnx', 'F6': ' ERC20 uniq sent addr.1', 'F8': 'avg val received', 'F1': 'Unique Sent To Addresses', 'F12': 'max value received ', 'F35': 'max val sent', 'F2': 'min val sent', 'F30': 'Number of Created Contracts', 'F25': 'total ether received', 'F11': ' ERC20 uniq sent addr', 'F32': ' ERC20 total Ether received', 'F23': 'Received Tnx', 'F17': ' ERC20 avg val sent', 'F38': 'total Ether sent', 'F34': ' ERC20 min val sent', 'F29': 'max val sent to contract', 'F33': 'total ether balance', 'F16': ' ERC20 max val sent', 'F3': ' Total ERC20 tnxs', 'F4': ' ERC20 total ether sent', 'F27': ' ERC20 avg val rec', 'F19': 'avg value sent to contract', 'F7': ' ERC20 min val rec', 'F37': ' ERC20 max val rec', 'F20': ' ERC20 total Ether sent contract', 'F22': 'min value sent to contract', 'F26': 'total ether sent contracts'}

{'F30': 'F9', 'F38': 'F24', 'F9': 'F18', 'F3': 'F15', 'F14': 'F21', 'F37': 'F28', 'F4': 'F13', 'F2': 'F36', 'F7': 'F10', 'F28': 'F5', 'F18': 'F14', 'F1': 'F31', 'F29': 'F6', 'F11': 'F8', 'F8': 'F1', 'F10': 'F12', 'F13': 'F35', 'F12': 'F2', 'F6': 'F30', 'F20': 'F25', 'F27': 'F11', 'F24': 'F32', 'F5': 'F23', 'F36': 'F17', 'F19': 'F38', 'F34': 'F34', 'F16': 'F29', 'F22': 'F33', 'F35': 'F16', 'F23': 'F3', 'F25': 'F4', 'F33': 'F27', 'F17': 'F19', 'F31': 'F7', 'F32': 'F37', 'F26': 'F20', 'F15': 'F22', 'F21': 'F26'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

KNeighborsClassifier

C1

Credit Risk Classification

According to the machine learning model, it is more likely that the case's label is C1, with a certainty of 100.0%, and this prediction decision is mainly based on the effects of the following features: F5, F3, F11, F7, and F1 on the model. Apart from F1 and F7, all the other variables mentioned above have a strong positive influence, improving the odds of the prediction class, C1. Together with F1 and F7, the values of variables F10 and F8 indicate that C2 could be the correct label instead. Unlike the top positive variables, F5, F3, and F11, each of these negative variables has a moderate contribution to the final decision. The features F9, F4, F2, and F6 are shown to have made minor contributions to the model's decision in this case. In summary, with only the positive contributions from F5, F3, F11, F2, and F9, the model is very certain of the classification output as indicated by the predicted probabilities across C1 and C2.

{'C1': '100.00%', 'C2': '0.00%'}

{'F5': 'fea_4', 'F3': 'fea_8', 'F11': 'fea_2', 'F1': 'fea_9', 'F7': 'fea_6', 'F10': 'fea_10', 'F8': 'fea_1', 'F9': 'fea_7', 'F4': 'fea_11', 'F6': 'fea_3', 'F2': 'fea_5'}

{'F4': 'F5', 'F8': 'F3', 'F2': 'F11', 'F9': 'F1', 'F6': 'F7', 'F10': 'F10', 'F1': 'F8', 'F7': 'F9', 'F11': 'F4', 'F3': 'F6', 'F5': 'F2'}

{'C2': 'C1', 'C1': 'C2'}

Low

{'C1': 'Low', 'C2': 'High'}

SVMClassifier_poly

C1

Employee Attrition

The classification findings by the model for the case here are as follows: there is a 97.67% chance that C1 is the correct label hence only a marginally low chance of 2.33% that C1 is not the correct label but C2 is. From the above findings, it is valid to conclude that the right class for the given case is C1, and the model is very certain of this decision. The features with the most control and influence on the classification above are F22, F26, F17, F11, and F29 but the influence of the remaining features is either moderate or low or negligible. Some of the features with moderate impact include F13, F3, F6, and F7. Those with low influence are F20, F16, F14, F9, and F28. Finally, those with negligible impact are F15, F24, F1, F4, F30, F18, F19, F25, F21, and F27 since their values are shown to have no impact on the classification made by the model here. The top positive features increasing the prediction likelihood of class C1 are F22, F29, and F12. Conversely, the negative features decreasing the odds in favour of C2 are primarily F26, F6, and F17.

{'C1': '97.67%', 'C2': '2.33%'}

{'F22': 'OverTime', 'F26': 'JobSatisfaction', 'F17': 'BusinessTravel', 'F11': 'MaritalStatus', 'F29': 'EnvironmentSatisfaction', 'F13': 'Department', 'F3': 'Age', 'F6': 'YearsInCurrentRole', 'F7': 'TotalWorkingYears', 'F8': 'WorkLifeBalance', 'F12': 'JobLevel', 'F10': 'JobInvolvement', 'F5': 'EducationField', 'F2': 'JobRole', 'F23': 'MonthlyIncome', 'F20': 'PerformanceRating', 'F16': 'DistanceFromHome', 'F14': 'Education', 'F9': 'Gender', 'F28': 'YearsWithCurrManager', 'F15': 'PercentSalaryHike', 'F24': 'RelationshipSatisfaction', 'F1': 'MonthlyRate', 'F4': 'DailyRate', 'F30': 'YearsSinceLastPromotion', 'F18': 'HourlyRate', 'F19': 'YearsAtCompany', 'F25': 'TrainingTimesLastYear', 'F21': 'StockOptionLevel', 'F27': 'NumCompaniesWorked'}

{'F26': 'F22', 'F30': 'F26', 'F17': 'F17', 'F25': 'F11', 'F28': 'F29', 'F21': 'F13', 'F1': 'F3', 'F14': 'F6', 'F11': 'F7', 'F20': 'F8', 'F5': 'F12', 'F29': 'F10', 'F22': 'F5', 'F24': 'F2', 'F6': 'F23', 'F19': 'F20', 'F3': 'F16', 'F27': 'F14', 'F23': 'F9', 'F16': 'F28', 'F9': 'F15', 'F18': 'F24', 'F7': 'F1', 'F2': 'F4', 'F15': 'F30', 'F4': 'F18', 'F13': 'F19', 'F12': 'F25', 'F10': 'F21', 'F8': 'F27'}

{'C1': 'C1', 'C2': 'C2'}

Stay

{'C1': 'Leave', 'C2': 'Leave'}

LogisticRegression

C1

Flight Price-Range Classification

The model is very confident that C1 is the most probable class for the given case, with a probability of 90.48% which means that the other labels are very unlikely. F5 and F9 are the most important variables with respect to this classification verdict while all other variables are shown to have a medium or low impact. Fortunately, the top variables, F5 and F9, have the same direction of influence, increasing the likelihood of C1. Furthermore, while F7 and F2 push the model to predict C1, those pushing for the assignment of a different label are F8, F1, and F11. Finally, many features have a fairly small impact on the final prediction made by the model here, but F6, F1, and F12 have the least impact.

{'C1': '90.48%', 'C3': '9.51%', 'C2': '0.01%'}

{'F5': 'Total_Stops', 'F9': 'Airline', 'F2': 'Destination', 'F7': 'Arrival_hour', 'F8': 'Source', 'F4': 'Duration_hours', 'F3': 'Dep_hour', 'F11': 'Dep_minute', 'F10': 'Arrival_minute', 'F6': 'Journey_month', 'F1': 'Journey_day', 'F12': 'Duration_mins'}

{'F12': 'F5', 'F9': 'F9', 'F11': 'F2', 'F5': 'F7', 'F10': 'F8', 'F7': 'F4', 'F3': 'F3', 'F4': 'F11', 'F6': 'F10', 'F2': 'F6', 'F1': 'F1', 'F8': 'F12'}

{'C3': 'C1', 'C1': 'C3', 'C2': 'C2'}

Low

{'C1': 'Low', 'C3': 'Moderate', 'C2': 'High'}

SVC

C2

Water Quality Classification

Despite the reasonably high confidence in the assigned label, the prediction probabilities across the two classes indicate that C1 might be the correct label. F4, F7, F2, and F8 are the factors whose major contributions resulted in the labelling choice mentioned above. According to the analysis, the top two factors, F4 and F7, have a negative influence, leading the classifier to classify the data as C1 rather than C2. F1 is the only other negative variable with a moderate effect when compared to the other two negative variables. Nevertheless, there are several factors, F2, F8, F9, F5, F3, and F6, that favourably support and encourage the classifier to assign C2. All in all, the degree of uncertainty in this classification instance might be explained by just looking at the negative factors' rather strong pull on the classifier towards C1.

{'C1': '38.68%', 'C2': '61.32%'}

{'F4': 'Sulfate', 'F7': 'Hardness', 'F2': 'ph', 'F8': 'Conductivity', 'F1': 'Turbidity', 'F9': 'Chloramines', 'F5': 'Solids', 'F3': 'Trihalomethanes', 'F6': 'Organic_carbon'}

{'F5': 'F4', 'F2': 'F7', 'F1': 'F2', 'F6': 'F8', 'F9': 'F1', 'F4': 'F9', 'F3': 'F5', 'F8': 'F3', 'F7': 'F6'}

{'C2': 'C1', 'C1': 'C2'}

Portable

{'C1': 'Not Portable', 'C2': 'Portable'}

MLPClassifier

C2

Ethereum Fraud Detection

C1 has a probability estimate of only 6.80%, while that of C2 is 93.20%; consequently, the most likely class for the given case is C2. The important or relevant features considered by the classifier are F11, F16, F4, F15, F28, F14, F9, F32, F17, F21, F23, F33, F26, F18, F2, F5, F37, F3, F7, and F1. Not all input features are relevant when determining the appropriate label and these irrelevant features include F38, F13, and F30. Furthermore, F11 and F16 have a strong positive effect, increasing the odds in favour of C2. In contrast, the F4, F28, and F15 are the negative features, lowering the odds of C2. Comparing the attributions of F11, F14, and F16 features to those of the negative features mentioned above, it is not surprising that the classifier is convinced that C2 is the most likely label here.

{'C1': '6.80%', 'C2': '93.20%'}

{'F11': 'Unique Received From Addresses', 'F16': ' ERC20 total Ether sent contract', 'F4': 'total ether received', 'F15': 'Sent tnx', 'F28': 'Number of Created Contracts', 'F14': ' ERC20 uniq rec token name', 'F9': ' ERC20 uniq rec contract addr', 'F32': 'max value received ', 'F17': 'total transactions (including tnx to create contract', 'F21': ' ERC20 uniq sent addr.1', 'F23': ' ERC20 uniq sent addr', 'F33': 'Received Tnx', 'F26': 'avg val received', 'F18': ' ERC20 uniq rec addr', 'F2': 'avg val sent', 'F5': 'min value received', 'F37': 'Unique Sent To Addresses', 'F3': ' ERC20 uniq sent token name', 'F7': 'Avg min between received tnx', 'F1': 'Time Diff between first and last (Mins)', 'F38': ' ERC20 min val rec', 'F30': ' ERC20 max val rec', 'F13': ' ERC20 min val sent', 'F22': ' ERC20 max val sent', 'F20': ' ERC20 avg val sent', 'F27': ' ERC20 avg val rec', 'F19': ' Total ERC20 tnxs', 'F10': ' ERC20 total ether sent', 'F24': ' ERC20 total Ether received', 'F36': 'total ether balance', 'F12': 'total ether sent contracts', 'F25': 'total Ether sent', 'F29': 'avg value sent to contract', 'F6': 'max val sent to contract', 'F31': 'min value sent to contract', 'F35': 'max val sent', 'F8': 'min val sent', 'F34': 'Avg min between sent tnx'}

{'F7': 'F11', 'F26': 'F16', 'F20': 'F4', 'F4': 'F15', 'F6': 'F28', 'F38': 'F14', 'F30': 'F9', 'F10': 'F32', 'F18': 'F17', 'F29': 'F21', 'F27': 'F23', 'F5': 'F33', 'F11': 'F26', 'F28': 'F18', 'F14': 'F2', 'F9': 'F5', 'F8': 'F37', 'F37': 'F3', 'F2': 'F7', 'F3': 'F1', 'F31': 'F38', 'F32': 'F30', 'F34': 'F13', 'F35': 'F22', 'F36': 'F20', 'F33': 'F27', 'F23': 'F19', 'F25': 'F10', 'F24': 'F24', 'F22': 'F36', 'F21': 'F12', 'F19': 'F25', 'F17': 'F29', 'F16': 'F6', 'F15': 'F31', 'F13': 'F35', 'F12': 'F8', 'F1': 'F34'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

BernoulliNB

C1

German Credit Evaluation

The model is not 100% convinced that the correct label for the data under consideration is C1 since there is a 26.27% chance that labelling the data as C2 is correct. All the input variables are shown to have some degree of influence on the classification decision, with the most influential variables being F2, F5, and F3, whereas F4 and F1 are the least influential. The impact of F7, F6, F8, and F9 can be considered moderate compared to the F2, F5, and F3. The uncertainty surrounding the above classification can be blamed on the fact that the majority of input variables have values suggesting that C2 could be the appropriate label. The negative features that decrease the prediction likelihood of C1 are F2, F3, F8, and F9. However, given that the prediction probability is about 73.73%, it can be said that the influence of positive features, F5, F7, F6, and F4, is enough to swing the model's verdict in favour of C1.

{'C1': '73.73%', 'C2': '26.27%'}

{'F2': 'Saving accounts', 'F5': 'Sex', 'F3': 'Housing', 'F7': 'Purpose', 'F6': 'Checking account', 'F8': 'Job', 'F9': 'Duration', 'F4': 'Age', 'F1': 'Credit amount'}

{'F5': 'F2', 'F2': 'F5', 'F4': 'F3', 'F9': 'F7', 'F6': 'F6', 'F3': 'F8', 'F8': 'F9', 'F1': 'F4', 'F7': 'F1'}

{'C2': 'C1', 'C1': 'C2'}

Good Credit

{'C1': 'Good Credit', 'C2': 'Bad Credit'}

SVMClassifier_poly

C2

Employee Attrition

The model predicted class C2 with an 81.98% prediction likelihood. F26 had the largest impact, followed by F6, F25, F16, F21, F28, F8, F3, F12, F15, F2, F20, F24, F5, F22, F11, F4, F1, F9, and finally, F13, which had the smallest non-zero impact. F26, the feature with the largest impact, contributed against the direction of the prediction, whereas F6, F25, F16, and F21 all contributed positively towards the prediction. Other features that had a negative influence on the prediction included F8 and F3, whereas F28 had a positive influence on the prediction. F27, F7, F18, and F29 are shown to have close to zero attribution in the model's prediction verdict in the given case.

{'C2': '81.98%', 'C1': '18.02%'}

{'F26': 'OverTime', 'F6': 'JobSatisfaction', 'F25': 'MaritalStatus', 'F16': 'Department', 'F21': 'NumCompaniesWorked', 'F28': 'BusinessTravel', 'F8': 'JobRole', 'F3': 'EnvironmentSatisfaction', 'F12': 'YearsInCurrentRole', 'F15': 'JobInvolvement', 'F2': 'WorkLifeBalance', 'F20': 'YearsSinceLastPromotion', 'F24': 'TotalWorkingYears', 'F5': 'JobLevel', 'F22': 'Age', 'F11': 'EducationField', 'F4': 'PerformanceRating', 'F1': 'MonthlyRate', 'F9': 'Education', 'F13': 'MonthlyIncome', 'F27': 'DailyRate', 'F7': 'YearsAtCompany', 'F18': 'RelationshipSatisfaction', 'F29': 'TrainingTimesLastYear', 'F14': 'StockOptionLevel', 'F19': 'Gender', 'F10': 'PercentSalaryHike', 'F30': 'HourlyRate', 'F17': 'DistanceFromHome', 'F23': 'YearsWithCurrManager'}

{'F26': 'F26', 'F30': 'F6', 'F25': 'F25', 'F21': 'F16', 'F8': 'F21', 'F17': 'F28', 'F24': 'F8', 'F28': 'F3', 'F14': 'F12', 'F29': 'F15', 'F20': 'F2', 'F15': 'F20', 'F11': 'F24', 'F5': 'F5', 'F1': 'F22', 'F22': 'F11', 'F19': 'F4', 'F7': 'F1', 'F27': 'F9', 'F6': 'F13', 'F2': 'F27', 'F13': 'F7', 'F18': 'F18', 'F12': 'F29', 'F10': 'F14', 'F23': 'F19', 'F9': 'F10', 'F4': 'F30', 'F3': 'F17', 'F16': 'F23'}

{'C1': 'C2', 'C2': 'C1'}

Stay

{'C2': 'Leave', 'C1': 'Leave'}

SVC

C1

German Credit Evaluation

This case's label has a 70.83 percent chance of being C1 and per the predicted likelihoods across the alternative labels, C3 has a 29.71 percent chance of being the correct label, however, the model is certain that C2 is not the true label. The most important variables are F5, F4, F9, and F1, whereas the remaining influential variables are listed in order of the magnitude of their contributions: F7, F2, F3, F6, and F8. Three of the nine variables have values that push towards the prediction of label C3 while the other attributes are referred to as positive since their values inspire the prediction of class C1. F5, F4, and F9 are the three attributes that have a negative influence on the prediction judgement, pushing it away from C1 towards the label C3. Finally, it is essential to highlight that the cumulative effect of positive attributes is greater than that of negative attributes, F9, F4, and F5.

{'C1': '70.83%', 'C3': '29.17%', 'C2': '0.0%'}

{'F1': 'Checking account', 'F5': 'Duration', 'F4': 'Housing', 'F9': 'Saving accounts', 'F7': 'Sex', 'F2': 'Age', 'F3': 'Purpose', 'F6': 'Job', 'F8': 'Credit amount'}

{'F6': 'F1', 'F8': 'F5', 'F4': 'F4', 'F5': 'F9', 'F2': 'F7', 'F1': 'F2', 'F9': 'F3', 'F3': 'F6', 'F7': 'F8'}

{'C3': 'C1', 'C1': 'C3', 'C2': 'C2'}

Good Credit

{'C1': 'Good Credit', 'C3': 'Bad Credit', 'C2': 'Other'}

SVC

C1

Vehicle Insurance Claims

First of all, the classification decision is solely based on the information or data supplied to the prediction model. According to the model, there is a 61.61% chance that C1 is the true label, and a 38.39% chance that C2 is the true label. Since the predicted probability of C1 is higher than that of C2, it is valid to conclude that C1 is most likely the true label. The main feature responsible for this classification is F26, with a very strong positive influence, driving the model's decision higher towards C1. The next set of relevant features are F13, F9, F27, F4, F17, F20, F16, and F10. Among all the features mentioned above, F13, F27, F4, F20, and F16 have negative contributions that are responsible for the decrease in the probability that C1 is the true label. This implies that the contributions of F9, F17, and F10 combined with that of F26 explain why the model is moderately certain that C1 is the true label.

{'C1': '61.61%', 'C2': '38.39%'}

{'F26': 'incident_severity', 'F13': 'insured_hobbies', 'F9': 'authorities_contacted', 'F27': 'insured_education_level', 'F4': 'umbrella_limit', 'F17': 'insured_relationship', 'F20': 'auto_make', 'F16': 'insured_occupation', 'F10': 'capital-gains', 'F11': 'policy_deductable', 'F31': 'policy_state', 'F25': 'auto_year', 'F19': 'insured_sex', 'F22': 'vehicle_claim', 'F30': 'incident_city', 'F21': 'number_of_vehicles_involved', 'F6': 'insured_zip', 'F2': 'injury_claim', 'F8': 'property_claim', 'F1': 'incident_type', 'F12': 'total_claim_amount', 'F33': 'police_report_available', 'F3': 'property_damage', 'F23': 'incident_state', 'F15': 'policy_annual_premium', 'F14': 'incident_hour_of_the_day', 'F32': 'collision_type', 'F7': 'capital-loss', 'F24': 'bodily_injuries', 'F29': 'policy_csl', 'F18': 'witnesses', 'F28': 'age', 'F5': 'months_as_customer'}

{'F27': 'F26', 'F23': 'F13', 'F28': 'F9', 'F21': 'F27', 'F5': 'F4', 'F24': 'F17', 'F33': 'F20', 'F22': 'F16', 'F7': 'F10', 'F3': 'F11', 'F18': 'F31', 'F17': 'F25', 'F20': 'F19', 'F16': 'F22', 'F30': 'F30', 'F10': 'F21', 'F6': 'F6', 'F14': 'F2', 'F15': 'F8', 'F25': 'F1', 'F13': 'F12', 'F32': 'F33', 'F31': 'F3', 'F29': 'F23', 'F4': 'F15', 'F9': 'F14', 'F26': 'F32', 'F8': 'F7', 'F11': 'F24', 'F19': 'F29', 'F12': 'F18', 'F2': 'F28', 'F1': 'F5'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

GradientBoostingClassifier

C2

Paris House Classification

Because the prediction probability of C1 is barely 0.70 percent, the classifier outputs the label C2 with near 100 percent confidence based on the values of the input attributes. The effects of F13, F10, and F14 on the aforementioned classification decision are significant. The values of these features are given greater emphasis by the classifier than the others. F14 is has a negative impact among these top features, pushing the prediction judgement towards the least likely class, C1 whereas on the other hand, F13 and F10 are referred to as positive features since they improve the likelihood of the C2 label rather than the C1 label. Finally, unlike the others, the values of F2, F17, F3, and F1 have only a little influence on the label selection made here.

{'C2': '99.30%', 'C1': '0.70%'}

{'F13': 'isNewBuilt', 'F14': 'hasYard', 'F10': 'hasPool', 'F9': 'hasStormProtector', 'F11': 'made', 'F7': 'hasGuestRoom', 'F16': 'squareMeters', 'F12': 'floors', 'F15': 'cityCode', 'F6': 'basement', 'F4': 'price', 'F8': 'numPrevOwners', 'F5': 'numberOfRooms', 'F2': 'attic', 'F17': 'cityPartRange', 'F3': 'garage', 'F1': 'hasStorageRoom'}

{'F3': 'F13', 'F1': 'F14', 'F2': 'F10', 'F4': 'F9', 'F12': 'F11', 'F16': 'F7', 'F6': 'F16', 'F8': 'F12', 'F9': 'F15', 'F13': 'F6', 'F17': 'F4', 'F11': 'F8', 'F7': 'F5', 'F14': 'F2', 'F10': 'F17', 'F15': 'F3', 'F5': 'F1'}

{'C1': 'C2', 'C2': 'C1'}

Basic

{'C2': 'Basic', 'C1': 'Luxury'}

SGDClassifier

C3

Flight Price-Range Classification

The classification algorithm arrived at the prediction output based on the variables or information supplied about the case under consideration. The prediction probabilities across the three-class labels, C2, C3, and C1, respectively, are 28.17%, 50.21%, and 21.62%, making C3 the label assigned by the algorithm, judged based on the prediction probabilities. The attributions analysis suggests that F8, F2, F10, and F1 are the positive features that increase the algorithm's prediction response in favour of C3. On the other hand, F3, F9, F5, F4, F12, F11, F6, and F7 have negative contributions in support of labelling the case as either C2 or C1. Overall, judging by the degree of contributions of the positive features, it is not surprising that the algorithm is moderately certain that neither C2 nor C1 is the most probable label for the case under consideration here.

{'C2': '28.17%', 'C3': '50.21%', 'C1': '21.62%'}

{'F8': 'Airline', 'F2': 'Total_Stops', 'F10': 'Arrival_minute', 'F3': 'Journey_day', 'F1': 'Dep_hour', 'F9': 'Source', 'F5': 'Dep_minute', 'F4': 'Duration_hours', 'F12': 'Destination', 'F11': 'Journey_month', 'F6': 'Duration_mins', 'F7': 'Arrival_hour'}

{'F9': 'F8', 'F12': 'F2', 'F6': 'F10', 'F1': 'F3', 'F3': 'F1', 'F10': 'F9', 'F4': 'F5', 'F7': 'F4', 'F11': 'F12', 'F2': 'F11', 'F8': 'F6', 'F5': 'F7'}

{'C2': 'C2', 'C3': 'C3', 'C1': 'C1'}

Moderate

{'C2': 'Low', 'C3': 'Moderate', 'C1': 'High'}

RandomForestClassifier

C2

Paris House Classification

Judging based on the information provided on the case under consideration, the model outputs that the prediction probability of C1 is only 0.48%, indicating that with about 99.52% certainty, the true label here is C2 and in simple terms, the model is very confident that the true label for the case under consideration is C2. The higher degree of certainty in the above classification can be attributed solely to the positive contributions of influential features F17, F13, and F10. Analysis indicates that all the remaining features such as F1, F3, F11, F16, and F7 have moderate to low contributions towards the prediction conclusions above, whereas F4, F14, F5, and F12 are the least relevant features here. The very marginal decrease in the C2's prediction likelihood could be attributed to the influence of negative features F3, F16, F6, F12, and F14 since their contributions support labelling the case as C1 instead. Moderate positive features further driving the model to label this case as C2 are F1, F11, F9, and F7.

{'C2': '99.52%', 'C1': '0.48%'}

{'F17': 'isNewBuilt', 'F13': 'hasYard', 'F10': 'hasPool', 'F1': 'made', 'F3': 'hasStormProtector', 'F11': 'hasGuestRoom', 'F16': 'squareMeters', 'F7': 'floors', 'F9': 'price', 'F8': 'cityCode', 'F15': 'basement', 'F6': 'numPrevOwners', 'F2': 'cityPartRange', 'F4': 'numberOfRooms', 'F14': 'attic', 'F5': 'garage', 'F12': 'hasStorageRoom'}

{'F3': 'F17', 'F1': 'F13', 'F2': 'F10', 'F12': 'F1', 'F4': 'F3', 'F16': 'F11', 'F6': 'F16', 'F8': 'F7', 'F17': 'F9', 'F9': 'F8', 'F13': 'F15', 'F11': 'F6', 'F10': 'F2', 'F7': 'F4', 'F14': 'F14', 'F15': 'F5', 'F5': 'F12'}

{'C1': 'C2', 'C2': 'C1'}

Basic

{'C2': 'Basic', 'C1': 'Luxury'}

GradientBoostingClassifier

C2

Basketball Players Career Length Prediction

Judging based on the values of the variables passed to the model with respect to the case under consideration, the output labelling decision is as follows: there is about an 83.98% chance that C2 is the correct label, whereas the likelihood of C1 is only 16.02%, hence the label choice with a higher confidence level is C2. The top-variables influencing this decision are F13, F1, F8, and F5, while the least important variables are F18, F4, and F2. According to the variable contributions analysis performed, only the input variables F14, F12, F11, and F6 exhibit negative attributions, pushing the prediction decision towards the alternative label, C1. The other variables positively support the C2 prediction, shifting the verdict strongly away from the C1 class. In conclusion, positive variables such as F13, F1, F8, F5, F15, and F9 have a higher joint contribution compared to the negative features, which can explain why the model is certain that C2 is the most probable label.

{'C2': '83.98%', 'C1': '16.02%'}

{'F13': 'GamesPlayed', 'F1': 'OffensiveRebounds', 'F8': 'FreeThrowPercent', 'F5': 'FieldGoalPercent', 'F19': '3PointPercent', 'F14': '3PointAttempt', 'F9': 'FieldGoalsMade', 'F15': 'Blocks', 'F3': 'DefensiveRebounds', 'F17': 'Turnovers', 'F12': 'Rebounds', 'F11': 'MinutesPlayed', 'F10': 'FreeThrowAttempt', 'F16': 'Assists', 'F6': '3PointMade', 'F7': 'FieldGoalsAttempt', 'F18': 'PointsPerGame', 'F2': 'Steals', 'F4': 'FreeThrowMade'}

{'F1': 'F13', 'F13': 'F1', 'F12': 'F8', 'F6': 'F5', 'F9': 'F19', 'F8': 'F14', 'F4': 'F9', 'F18': 'F15', 'F14': 'F3', 'F19': 'F17', 'F15': 'F12', 'F2': 'F11', 'F11': 'F10', 'F16': 'F16', 'F7': 'F6', 'F5': 'F7', 'F3': 'F18', 'F17': 'F2', 'F10': 'F4'}

{'C2': 'C2', 'C1': 'C1'}

More than 5

{'C2': 'More than 5', 'C1': 'Less than 5'}

RandomForestClassifier

C3

Mobile Price-Range Classification

The model predicts the class label C3 for the given test instance with a likelihood of about 69.23%. However, there is about a 30.77% chance that the true class label is C2, while the others, C1 and C4, have a 0.0% likelihood. The top features contributing to this prediction decision are F8, F14, F12, and F6, whereas the least important are F3, F11, and F20. Among the top features, while F8 and F14 have values that shift the prediction decision towards the C3 class label, the values of F12 and F6 suggest that the true label could likely be C2. For the features with moderate influence on the decision, F16, F5, F9, and F4 have negative contributions, further decreasing the confidence level in the C3 assignment. On the other hand, the moderate positive influences of F19, F15, F1, F7, and F13 drive the decision further towards the C3 label. Considering the attributions of the input features, it is surprising that the confidence level is just 69.23% since the top feature, F8, has the highest contribution among all the input features. Finally, the values of F2, F11, and F20, though shown to be less important when deciding the correct label for the given case, have positive contributions to the prediction with respect to the given case.

{'C1': '0.00%', 'C3': '69.23%', 'C2': '30.77%', 'C4': '0.00%'}

{'F8': 'ram', 'F14': 'touch_screen', 'F12': 'int_memory', 'F6': 'battery_power', 'F19': 'mobile_wt', 'F15': 'sc_w', 'F16': 'four_g', 'F7': 'talk_time', 'F1': 'sc_h', 'F13': 'wifi', 'F5': 'fc', 'F9': 'three_g', 'F4': 'dual_sim', 'F18': 'n_cores', 'F10': 'px_height', 'F17': 'blue', 'F2': 'clock_speed', 'F3': 'px_width', 'F11': 'm_dep', 'F20': 'pc'}

{'F11': 'F8', 'F19': 'F14', 'F4': 'F12', 'F1': 'F6', 'F6': 'F19', 'F13': 'F15', 'F17': 'F16', 'F14': 'F7', 'F12': 'F1', 'F20': 'F13', 'F3': 'F5', 'F18': 'F9', 'F16': 'F4', 'F7': 'F18', 'F9': 'F10', 'F15': 'F17', 'F2': 'F2', 'F10': 'F3', 'F5': 'F11', 'F8': 'F20'}

{'C3': 'C1', 'C4': 'C3', 'C2': 'C2', 'C1': 'C4'}

r2

{'C1': 'r1', 'C3': 'r2', 'C2': 'r3', 'C4': 'r4'}

KNeighborsClassifier

C1

Water Quality Classification

The given case is likely C1 with a confidence level of 87.50% judged based on the values of the input features supplied to the classifier and according to the attributions analysis, F1 and F5 have a high degree of impact. F8, F6, F2, F4, and F9 have a moderate degree of impact while on the contrary F7 and F3 have little impact. Examining further, the values of F1, F5, F8, and F6 all have a positive influence on the classifier supporting the label assignment decision for the given test case. F2 and F9 are also positively supporting features, whereas F4 has a negative influence on the final classification. Finally, F7 and F3 both have very little contributions, though F3 has significantly less than even F7.

{'C1': '87.50%', 'C2': '12.50%'}

{'F1': 'Hardness', 'F5': 'Sulfate', 'F8': 'Solids', 'F6': 'ph', 'F2': 'Organic_carbon', 'F9': 'Conductivity', 'F4': 'Trihalomethanes', 'F7': 'Turbidity', 'F3': 'Chloramines'}

{'F2': 'F1', 'F5': 'F5', 'F3': 'F8', 'F1': 'F6', 'F7': 'F2', 'F6': 'F9', 'F8': 'F4', 'F9': 'F7', 'F4': 'F3'}

{'C2': 'C1', 'C1': 'C2'}

Not Portable

{'C1': 'Not Portable', 'C2': 'Portable'}

RandomForestClassifier

C3

Mobile Price-Range Classification

The label for this example is estimated to be C3 among the four possible classes, with a 73.08 percent chance of being true. C2 is the next most likely label, with a probability of roughly 26.92 percent. The above prediction assessment is mostly dependent on the values of the variables F18, F7, F14, F19, and F8. F18 had the greatest influence, followed by F14, F7, F8, and F19. The positive variables F18, F7, F6, and F16 outnumber the negative variables F14, F8, F19, and F2. Twelve of the twenty variables have values that tilt the prediction towards one of the three other probable classifications. As a result, it is not unexpected that the model is not completely certain of the C3 assigned. Given that the chance of C3's being accurate is 73.08 percent, the model appears to be relatively confident in its final judgement for the data instance under review.

{'C3': '73.08%', 'C2': '26.92%', 'C4': '0.00%', 'C1': '0.00%'}

{'F18': 'ram', 'F14': 'px_width', 'F7': 'battery_power', 'F8': 'px_height', 'F19': 'n_cores', 'F6': 'dual_sim', 'F16': 'touch_screen', 'F2': 'int_memory', 'F9': 'wifi', 'F1': 'fc', 'F13': 'four_g', 'F3': 'm_dep', 'F11': 'pc', 'F5': 'mobile_wt', 'F15': 'talk_time', 'F17': 'three_g', 'F12': 'sc_h', 'F10': 'sc_w', 'F20': 'blue', 'F4': 'clock_speed'}

{'F11': 'F18', 'F10': 'F14', 'F1': 'F7', 'F9': 'F8', 'F7': 'F19', 'F16': 'F6', 'F19': 'F16', 'F4': 'F2', 'F20': 'F9', 'F3': 'F1', 'F17': 'F13', 'F5': 'F3', 'F8': 'F11', 'F6': 'F5', 'F14': 'F15', 'F18': 'F17', 'F12': 'F12', 'F13': 'F10', 'F15': 'F20', 'F2': 'F4'}

{'C2': 'C3', 'C1': 'C2', 'C4': 'C4', 'C3': 'C1'}

r1

{'C3': 'r1', 'C2': 'r2', 'C4': 'r3', 'C1': 'r4'}

BernoulliNB

C1

Personal Loan Modelling

The model has classified the instance as C1 due to the effects of the following features: F2, F7, F4, and F6. Based on the values of these variables, the likelihood of the C1 label is 65.51 percent. F6 and F4 are the top positively contributing variables, whereas F2 and F7 are the most adversely contributing variables. Unlike F6 and F4, which have greater influences on the model's prediction choice in this situation, F3 and F9 have fairly modest positive influences. Finally, F8, F5, and F1 show negative predictive effects, however, as compared to F2, their attributions are modest.

{'C2': '34.49%', 'C1': '65.51%'}

{'F6': 'CD Account', 'F4': 'Income', 'F2': 'CCAvg', 'F7': 'Securities Account', 'F3': 'Education', 'F9': 'Mortgage', 'F8': 'Age', 'F5': 'Family', 'F1': 'Extra_service'}

{'F8': 'F6', 'F2': 'F4', 'F4': 'F2', 'F7': 'F7', 'F5': 'F3', 'F6': 'F9', 'F1': 'F8', 'F3': 'F5', 'F9': 'F1'}

{'C1': 'C2', 'C2': 'C1'}

Accept

{'C2': 'Reject', 'C1': 'Accept'}

DecisionTreeClassifier

C1

Insurance Churn

Considering the predicted likelihoods across the classes, C1 is confidently chosen as the true label since its likelihood is 93.27%, implying that the likelihood of C2 is only about 6.73%. F13 and F3 are the two features with a very strong positive influence, favouring the prediction of class C1. The following features have a moderate effect and are listed in descending order of influence: F7 and F10 have a negative effect, while F11 and F5 have a positive effect on the prediction of C1. Similar to F7 and F10, the features F8 and F9 also negatively affected the prediction decision. Finally, the values of F14, F15, F1, and F16 are the least important to the model decision for this case.

{'C2': '6.73%', 'C1': '93.27%'}

{'F13': 'feature15', 'F3': 'feature14', 'F7': 'feature10', 'F10': 'feature11', 'F11': 'feature5', 'F5': 'feature13', 'F8': 'feature4', 'F9': 'feature3', 'F2': 'feature12', 'F12': 'feature1', 'F6': 'feature7', 'F4': 'feature2', 'F14': 'feature6', 'F15': 'feature0', 'F1': 'feature9', 'F16': 'feature8'}

{'F9': 'F13', 'F8': 'F3', 'F4': 'F7', 'F5': 'F10', 'F15': 'F11', 'F7': 'F5', 'F14': 'F8', 'F13': 'F9', 'F6': 'F2', 'F11': 'F12', 'F1': 'F6', 'F12': 'F4', 'F16': 'F14', 'F10': 'F15', 'F3': 'F1', 'F2': 'F16'}

{'C1': 'C2', 'C2': 'C1'}

Leave

{'C2': 'Stay', 'C1': 'Leave'}

GradientBoostingClassifier

C1

Basketball Players Career Length Prediction

The classification output is C1, however, the classifier is somewhat unsure about this prediction decision because the corresponding predicted probability is only 55.19%. F3 is by far the most influential feature whereas F1, F13, and F14 have been recognised as having the biggest effect on prediction output here after F3. The combination of F3, F1, F13, F14, and F10 features has resulted in the classification choice being altered from C1 to C2. While F9, F7, and F2 all have a minor influence on the classification, F9 is the only one that has a positive impact on the C1 classification. In this case, many features had lower influence on the prediction, with F19, F8, F12, F18, and F16 having a marginal effect.

{'C2': '44.81%', 'C1': '55.19%'}

{'F3': 'GamesPlayed', 'F1': 'OffensiveRebounds', 'F13': 'FieldGoalPercent', 'F14': 'FreeThrowPercent', 'F10': '3PointPercent', 'F9': '3PointAttempt', 'F7': 'FieldGoalsMade', 'F2': 'Blocks', 'F5': 'DefensiveRebounds', 'F11': 'Turnovers', 'F15': 'Rebounds', 'F17': 'MinutesPlayed', 'F4': 'FreeThrowAttempt', 'F6': '3PointMade', 'F19': 'Assists', 'F8': 'PointsPerGame', 'F12': 'FreeThrowMade', 'F18': 'FieldGoalsAttempt', 'F16': 'Steals'}

{'F1': 'F3', 'F13': 'F1', 'F6': 'F13', 'F12': 'F14', 'F9': 'F10', 'F8': 'F9', 'F4': 'F7', 'F18': 'F2', 'F14': 'F5', 'F19': 'F11', 'F15': 'F15', 'F2': 'F17', 'F11': 'F4', 'F7': 'F6', 'F16': 'F19', 'F3': 'F8', 'F10': 'F12', 'F5': 'F18', 'F17': 'F16'}

{'C2': 'C2', 'C1': 'C1'}

Less than 5

{'C2': 'More than 5', 'C1': 'Less than 5'}

LogisticRegression

C1

Customer Churn Modelling

Judging based on the values of the input variables, the classification algorithm labels the case as C1 since its prediction likelihood is equal to 88.69%. The prediction decision is primarily based on the contributions of F7, F4, and F10, however, F6, F1, and F9 are shown to be the least important variables. Regarding the direction of influence of the variables, F7, F10, F2, F6, and F1 are the positive variables that increase the odds of C1 being the correct label. Driving the prediction toward the alternative label, C2, are the variables F4, F3, F8, F5, and F9. Owing to the fact that the most influential variables, F7 and F10, have strong positive attributions, outweighing the contributions of the negative variables, it is not surprising that the algorithm is certain about the decision made.

{'C1': '88.69%', 'C2': '11.31%'}

{'F7': 'IsActiveMember', 'F10': 'NumOfProducts', 'F4': 'Geography', 'F3': 'Gender', 'F8': 'Age', 'F5': 'CreditScore', 'F2': 'EstimatedSalary', 'F6': 'Balance', 'F1': 'Tenure', 'F9': 'HasCrCard'}

{'F9': 'F7', 'F7': 'F10', 'F2': 'F4', 'F3': 'F3', 'F4': 'F8', 'F1': 'F5', 'F10': 'F2', 'F6': 'F6', 'F5': 'F1', 'F8': 'F9'}

{'C2': 'C1', 'C1': 'C2'}

Stay

{'C1': 'Stay', 'C2': 'Leave'}

BernoulliNB

C2

Water Quality Classification

The classification algorithm predicts class C2 with a confidence level of 61.55% and this implies that the probability of the alternative label is only 38.45%. In this case, the top features driving the prediction decision are F7, F8, F2, and F4, followed by F3, F5, F9, F6, and finally F1. Based on the inspections performed to understand the direction of influence of the input features, it can be concluded that F7 has the strongest positive contribution, while F2 has the strongest negative contribution and conversely, all the remaining features have moderate contributions. The other positive features are F8, F3, F5, and F6, whereas the remaining negatives are F4, F9, and F1. All things considered, the influence of the negative features indicates that the likelihood of the C1 label is 38.45% while the positive contributions push the prediction higher towards C2 resulting in the 61.55% prediction confidence.

{'C2': '61.55%', 'C1': '38.45%'}

{'F7': 'Sulfate', 'F8': 'ph', 'F2': 'Trihalomethanes', 'F4': 'Chloramines', 'F3': 'Organic_carbon', 'F5': 'Hardness', 'F9': 'Solids', 'F6': 'Turbidity', 'F1': 'Conductivity'}

{'F5': 'F7', 'F1': 'F8', 'F8': 'F2', 'F4': 'F4', 'F7': 'F3', 'F2': 'F5', 'F3': 'F9', 'F9': 'F6', 'F6': 'F1'}

{'C1': 'C2', 'C2': 'C1'}

Not Portable

{'C2': 'Not Portable', 'C1': 'Portable'}

RandomForestClassifier

C2

Flight Price-Range Classification

The classification model's decision about the true label for the case is based on the information provided to it. Among the three labels, C2, C1, and C3, the model shows without a doubt that neither C1 nor C3 is the true label, given that the probability of C2 being the true label is 100.0%. F4, F5, and F8 are the main contributing factors or variables in the final verdict here since their respective influence outranks the remaining variables. In fact, analysis indicates that F10, F1, and F12 are the least influential variables since they receive little emphasis from the model when making the labelling decision here. In between F4, F5, and F8, and F10, F1, F2, and F12, are the variables such as F9, F6, F3, and F7 with moderate influence on the classification decision here. Among the variables passed to the model, only F9, F7, and F10 are shown to have negative contributions, which suggests that perhaps the true label could be either of the remaining labels. However, given the 100.0% predicted likelihood of C2, it is reasonable to deduce that the positive variables, such as F4, F5, F8, F6, F11, and F3, significantly influence the model's judgement towards C2.

{'C2': '100.00%', 'C1': '0.00%', 'C3': '0.00%'}

{'F4': 'Duration_hours', 'F5': 'Airline', 'F8': 'Total_Stops', 'F9': 'Journey_day', 'F6': 'Source', 'F3': 'Duration_mins', 'F7': 'Arrival_hour', 'F11': 'Destination', 'F2': 'Arrival_minute', 'F10': 'Dep_minute', 'F1': 'Journey_month', 'F12': 'Dep_hour'}

{'F7': 'F4', 'F9': 'F5', 'F12': 'F8', 'F1': 'F9', 'F10': 'F6', 'F8': 'F3', 'F5': 'F7', 'F11': 'F11', 'F6': 'F2', 'F4': 'F10', 'F2': 'F1', 'F3': 'F12'}

{'C2': 'C2', 'C3': 'C1', 'C1': 'C3'}

Low

{'C2': 'Low', 'C1': 'Moderate', 'C3': 'High'}

LogisticRegression

C2

Basketball Players Career Length Prediction

According to the model, C2 is the class with the higher probability, which is equal to 52.57 percent, of being the label for this selected instance or case. Conversely, there is a 47.43 percent chance that C1 is the correct label showing that the model is less certain about the classification verdict in this case. This uncertainty can be linked to the fact that the majority of variables have values that favour assigning C1. The only variables increasing the model's response to prediction C2 are the positive variables namely: F7, F1, F14, F5, F16, F2, and F18. The top negative variables decreasing the likelihood of C2 are F19 and F11 supported by other negative variables, F10, F9, and F15, that further shift the verdict towards C1.

{'C1': '47.43%', 'C2': '52.57%'}

{'F19': '3PointMade', 'F7': '3PointAttempt', 'F11': 'FreeThrowMade', 'F1': 'FreeThrowAttempt', 'F10': 'GamesPlayed', 'F9': 'OffensiveRebounds', 'F15': 'FieldGoalsAttempt', 'F14': 'DefensiveRebounds', 'F6': 'Assists', 'F5': 'MinutesPlayed', 'F16': 'FieldGoalsMade', 'F12': 'Blocks', 'F3': 'Rebounds', 'F13': 'FieldGoalPercent', 'F2': 'Steals', 'F17': 'PointsPerGame', 'F4': 'FreeThrowPercent', 'F18': 'Turnovers', 'F8': '3PointPercent'}

{'F7': 'F19', 'F8': 'F7', 'F10': 'F11', 'F11': 'F1', 'F1': 'F10', 'F13': 'F9', 'F5': 'F15', 'F14': 'F14', 'F16': 'F6', 'F2': 'F5', 'F4': 'F16', 'F18': 'F12', 'F15': 'F3', 'F6': 'F13', 'F17': 'F2', 'F3': 'F17', 'F12': 'F4', 'F19': 'F18', 'F9': 'F8'}

{'C2': 'C1', 'C1': 'C2'}

Less than 5

{'C1': 'More than 5', 'C2': 'Less than 5'}

RandomForestClassifier

C2

Printer Sales

According to the predicted likelihoods across the classes, C1 has a 17.0% chance of being the true label for the given data or case, implying that C2 is the most likely label. F16, F14, and F12 are the most important factors that led to the classification judgments above. The remaining factors have a minor or non-existent impact on the classifier. The classifier most likely ignored the values of F2, F1, F11, F22, F8, and F15 when giving a label to this case since their relative degrees of impact are extremely near to zero. F3, F25, F19, F21, F13, and F5 are considered negative factors among the significant factors because their contributions to the choice tend to reduce the chance that C2 is the correct label. These negatives features lend themselves to the case being classified as C1 but the remaining features contribute positively, raising the likelihood of the C2 classification.

{'C2': '83.00%', 'C1': '17.00%'}

{'F16': 'X8', 'F14': 'X24', 'F12': 'X1', 'F10': 'X2', 'F23': 'X10', 'F7': 'X15', 'F3': 'X25', 'F4': 'X23', 'F6': 'X18', 'F25': 'X4', 'F19': 'X7', 'F9': 'X17', 'F20': 'X3', 'F18': 'X22', 'F21': 'X5', 'F13': 'X9', 'F5': 'X12', 'F26': 'X19', 'F17': 'X11', 'F24': 'X16', 'F2': 'X14', 'F1': 'X21', 'F11': 'X20', 'F22': 'X13', 'F8': 'X6', 'F15': 'X26'}

{'F8': 'F16', 'F24': 'F14', 'F1': 'F12', 'F2': 'F10', 'F10': 'F23', 'F15': 'F7', 'F25': 'F3', 'F23': 'F4', 'F18': 'F6', 'F4': 'F25', 'F7': 'F19', 'F17': 'F9', 'F3': 'F20', 'F22': 'F18', 'F5': 'F21', 'F9': 'F13', 'F12': 'F5', 'F19': 'F26', 'F11': 'F17', 'F16': 'F24', 'F14': 'F2', 'F21': 'F1', 'F20': 'F11', 'F13': 'F22', 'F6': 'F8', 'F26': 'F15'}

{'C2': 'C2', 'C1': 'C1'}

Less

{'C2': 'Less', 'C1': 'More'}

RandomForestClassifier

C1

Credit Risk Classification

According to the ML model, C1 is the most likely class label, and we can conclude that the model is quite confident about the decision given that the probability of having C2 as the correct label is only 7.0%. For the case under study, analysis indicates that F10, F4, F6, and F7 are essentially the negative set of features that push the forecast higher towards C2 instead of C1, while F2, F11, F9, and F3 increase the odds of the prediction being equal to C1. In general, the most relevant feature is F2, while F5 and F1 are the least relevant features, with marginal influence on the above classification verdict. In summary, given the very strong positive influence of F2 together with the moderate influence of the other positives, F11, F3, and F9, it is not strange that the model chose to label the case as C1 instead of C2.

{'C1': '93.00%', 'C2': '7.00%'}

{'F2': 'fea_4', 'F6': 'fea_10', 'F11': 'fea_8', 'F10': 'fea_7', 'F3': 'fea_2', 'F9': 'fea_3', 'F4': 'fea_5', 'F7': 'fea_1', 'F8': 'fea_9', 'F1': 'fea_6', 'F5': 'fea_11'}

{'F4': 'F2', 'F10': 'F6', 'F8': 'F11', 'F7': 'F10', 'F2': 'F3', 'F3': 'F9', 'F5': 'F4', 'F1': 'F7', 'F9': 'F8', 'F6': 'F1', 'F11': 'F5'}

{'C1': 'C1', 'C2': 'C2'}

Low

{'C1': 'Low', 'C2': 'High'}

MLPClassifier

C1

Annual Income Earnings

Because the confidence level associated with the other class, C2, is just 2.29%, the model predicts that the given example is likely C1 and to be specific, the model is quite certain that the right label for the given case is C1. All the features are shown to have some degree of influence on the decision above, with F9 and F10 being the least relevant features, while F12 and F11 are the top features. From the analysis performed to understand how each feature contributes to the above prediction assertion, only the features F7, F6, F14, F1, F8, and F10, have negative influences, shifting the prediction verdict towards C2. The remaining features all contribute positively, strongly shifting the prediction towards the assigned label which could explain the prediction confidence level associated with label C1. The most positive features are F11, F2, and F12 with stronger push in favour of the output label and they are supported by other positive features such as F5, F3, F13, and F4 have a moderate degree of influence.

{'C2': '2.29%', 'C1': '97.71%'}

{'F12': 'Capital Gain', 'F11': 'Marital Status', 'F7': 'Capital Loss', 'F2': 'Relationship', 'F6': 'Hours per week', 'F13': 'Education', 'F4': 'Country', 'F5': 'Age', 'F3': 'Occupation', 'F1': 'Sex', 'F14': 'Education-Num', 'F8': 'Workclass', 'F9': 'fnlwgt', 'F10': 'Race'}

{'F11': 'F12', 'F6': 'F11', 'F12': 'F7', 'F8': 'F2', 'F13': 'F6', 'F4': 'F13', 'F14': 'F4', 'F1': 'F5', 'F7': 'F3', 'F10': 'F1', 'F5': 'F14', 'F2': 'F8', 'F3': 'F9', 'F9': 'F10'}

{'C1': 'C2', 'C2': 'C1'}

Above 50K

{'C2': 'Under 50K', 'C1': 'Above 50K'}

KNNClassifier

C2

Car Acceptability Valuation

The classifier made the prediction here based on the information provided about the case under consideration, and according to the classifier, the prediction probabilities or likelihoods across the labels C2 and C1 are 100.0% and 0.0%, respectively. All the input features are shown to have different degrees of influence on the final decision here by the classifier. The most influential features are F2 and F3, with F6 and F5 ranked as the least contributing factors. The values of F1 and F4 suggest that perhaps the true label could be C1 since they are the negative features. However, considering the confidence in C2, it is valid to conclude that the joint influence or contribution to the classification of the negative features with respect to the given case is outmatched by the joint positive attribution of F2, F3, F6, and F5.

{'C2': '100.00%', 'C1': '0.00%'}

{'F2': 'persons', 'F3': 'safety', 'F4': 'lug_boot', 'F1': 'buying', 'F6': 'doors', 'F5': 'maint'}

{'F4': 'F2', 'F6': 'F3', 'F5': 'F4', 'F1': 'F1', 'F3': 'F6', 'F2': 'F5'}

{'C1': 'C2', 'C2': 'C1'}

Unacceptable

{'C2': 'Unacceptable', 'C1': 'Acceptable'}

LogisticRegression

C1

Real Estate Investment

For the selected case, the model assigns the label C1. The prediction probability distribution across the classes C2 and C1 is 2.40% and 97.60%, respectively. The most important features considered for this prediction are F6, F12, F18, and F1, while on the other hand, the least relevant features with little contributions to the decision based on the analysis are F3, F13, F8, and F16. The top positive features Increasing the likelihood of the prediction being made are F6, F12, and F1. Pushing the prediction towards the alternative class C2, the top negative features are F18, F2, and F4. F17, F11, F19, F5, and F15 are some of the features that have a moderate impact on the classification decision in this case.

{'C2': '2.40%', 'C1': '97.60%'}

{'F6': 'Feature7', 'F12': 'Feature4', 'F18': 'Feature2', 'F1': 'Feature14', 'F2': 'Feature15', 'F4': 'Feature8', 'F17': 'Feature20', 'F11': 'Feature1', 'F19': 'Feature17', 'F15': 'Feature3', 'F5': 'Feature16', 'F7': 'Feature18', 'F20': 'Feature10', 'F14': 'Feature5', 'F9': 'Feature6', 'F10': 'Feature12', 'F3': 'Feature19', 'F13': 'Feature13', 'F8': 'Feature9', 'F16': 'Feature11'}

{'F11': 'F6', 'F9': 'F12', 'F1': 'F18', 'F17': 'F1', 'F4': 'F2', 'F3': 'F4', 'F20': 'F17', 'F7': 'F11', 'F6': 'F19', 'F8': 'F15', 'F18': 'F5', 'F19': 'F7', 'F13': 'F20', 'F2': 'F14', 'F10': 'F9', 'F15': 'F10', 'F5': 'F3', 'F16': 'F13', 'F12': 'F8', 'F14': 'F16'}

{'C2': 'C2', 'C1': 'C1'}

Invest

{'C2': 'Ignore', 'C1': 'Invest'}

MLPClassifier

C1

Vehicle Insurance Claims

The given instance was labelled as C1 by the model based on the values of its features. The model is about 79.64% certain about this prediction decision, hence, there is a slight chance that the label could be C2. Among the different features, the ones with the most impact on the model are F6, F17, F25, F11, and F13. The most negative feature is F6, and it is significantly pushing the narrative toward the prediction of C2. From this, it is foreseeable that there is a chance that the true label could be C2 which is about 20.36%. The influence of F6 and F25 is somewhat counterbalanced by the values of the features F17, F11, and F13. Other attributes that shift the decision in favour of C2 are F33 and F24. F1 shifts the decision further in the direction of C1 and in addition, F14 supports the model's prediction while the values of F19 and F8 of the given test case contradict the model's decision, decreasing the likelihood of C1. Among the features not relevant to this prediction decision for this case are F10, F4, F3, and F32.

{'C1': '79.64%', 'C2': '20.36%'}

{'F6': 'incident_severity', 'F17': 'insured_hobbies', 'F25': 'insured_relationship', 'F11': 'umbrella_limit', 'F13': 'insured_education_level', 'F33': 'authorities_contacted', 'F24': 'incident_type', 'F1': 'policy_csl', 'F14': 'number_of_vehicles_involved', 'F19': 'capital-loss', 'F8': 'property_damage', 'F21': 'insured_occupation', 'F22': 'age', 'F7': 'incident_state', 'F31': 'insured_zip', 'F2': 'collision_type', 'F12': 'property_claim', 'F29': 'injury_claim', 'F30': 'capital-gains', 'F16': 'witnesses', 'F10': 'incident_city', 'F4': 'police_report_available', 'F32': 'months_as_customer', 'F3': 'auto_year', 'F18': 'insured_sex', 'F27': 'policy_state', 'F23': 'vehicle_claim', 'F20': 'total_claim_amount', 'F26': 'bodily_injuries', 'F5': 'incident_hour_of_the_day', 'F28': 'policy_annual_premium', 'F15': 'policy_deductable', 'F9': 'auto_make'}

{'F27': 'F6', 'F23': 'F17', 'F24': 'F25', 'F5': 'F11', 'F21': 'F13', 'F28': 'F33', 'F25': 'F24', 'F19': 'F1', 'F10': 'F14', 'F8': 'F19', 'F31': 'F8', 'F22': 'F21', 'F2': 'F22', 'F29': 'F7', 'F6': 'F31', 'F26': 'F2', 'F15': 'F12', 'F14': 'F29', 'F7': 'F30', 'F12': 'F16', 'F30': 'F10', 'F32': 'F4', 'F1': 'F32', 'F17': 'F3', 'F20': 'F18', 'F18': 'F27', 'F16': 'F23', 'F13': 'F20', 'F11': 'F26', 'F9': 'F5', 'F4': 'F28', 'F3': 'F15', 'F33': 'F9'}

{'C1': 'C1', 'C2': 'C2'}

Not Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

RandomForestClassifier

C1

Ethereum Fraud Detection

According to the classification algorithm, the best label for the given case is C1, because there is little to no chance that C2 is the correct label. Not all of the features are found to contribute to the label given here. The following significant features are ordered in order of their effect on the algorithm's output: F15, F8, F29, F26, F13, F28, F18, F33, F19, F22, F24, F25, F21, F6, F37, F32, F9, F11, F2, F3. F17, F12, and F31, on the other hand, are unimportant features since they have almost no influence. Among the most influential features F15, F8, F29, F26, and F13, F29 is considered the most negative, dragging the verdict in a different direction, while the others have positive contributions, increasing the possibility that C1 is correct in this case. F18 is recognised as a positive feature with modest effect, whereas F28 and F33 are identified as negative features. Given that the majority of the top five attributes have positive contributions, boosting the likelihood that C1 is the correct label, it is not unexpected that the algorithm is quite confident in the assigned label's accuracy.

{'C2': '0.00%', 'C1': '100.00%'}

{'F15': ' ERC20 total Ether sent contract', 'F8': ' ERC20 min val rec', 'F29': 'total transactions (including tnx to create contract', 'F26': ' ERC20 max val rec', 'F13': ' Total ERC20 tnxs', 'F28': ' ERC20 uniq rec addr', 'F18': 'min val sent', 'F33': 'Time Diff between first and last (Mins)', 'F19': 'Sent tnx', 'F22': 'Avg min between received tnx', 'F24': 'min value received', 'F25': ' ERC20 total ether sent', 'F21': 'avg val sent', 'F6': 'max val sent', 'F37': 'Avg min between sent tnx', 'F32': 'Received Tnx', 'F9': ' ERC20 uniq sent token name', 'F11': 'Unique Sent To Addresses', 'F2': ' ERC20 uniq rec token name', 'F3': ' ERC20 uniq rec contract addr', 'F17': 'total Ether sent', 'F12': 'Number of Created Contracts', 'F31': ' ERC20 avg val sent', 'F30': ' ERC20 max val sent', 'F36': ' ERC20 min val sent', 'F7': ' ERC20 avg val rec', 'F1': 'Unique Received From Addresses', 'F27': 'max value received ', 'F14': ' ERC20 uniq sent addr.1', 'F16': 'total ether sent contracts', 'F4': 'avg val received', 'F20': ' ERC20 uniq sent addr', 'F5': 'min value sent to contract', 'F34': 'max val sent to contract', 'F10': ' ERC20 total Ether received', 'F23': 'avg value sent to contract', 'F38': 'total ether balance', 'F35': 'total ether received'}

{'F26': 'F15', 'F31': 'F8', 'F18': 'F29', 'F32': 'F26', 'F23': 'F13', 'F28': 'F28', 'F12': 'F18', 'F3': 'F33', 'F4': 'F19', 'F2': 'F22', 'F9': 'F24', 'F25': 'F25', 'F14': 'F21', 'F13': 'F6', 'F1': 'F37', 'F5': 'F32', 'F37': 'F9', 'F8': 'F11', 'F38': 'F2', 'F30': 'F3', 'F19': 'F17', 'F6': 'F12', 'F36': 'F31', 'F35': 'F30', 'F34': 'F36', 'F33': 'F7', 'F7': 'F1', 'F10': 'F27', 'F29': 'F14', 'F21': 'F16', 'F11': 'F4', 'F27': 'F20', 'F15': 'F5', 'F16': 'F34', 'F24': 'F10', 'F17': 'F23', 'F22': 'F38', 'F20': 'F35'}

{'C1': 'C2', 'C2': 'C1'}

Fraud

{'C2': 'Not Fraud', 'C1': 'Fraud'}

BernoulliNB

C1

Hotel Satisfaction

The classifier labbelled the given case as C1 with a confidence level of 98.89%, implying that the chance of C2 being the correct label is only about 1.11%. The classification output decision is solely based on the information supplied to the classifier about the case under review. We can rank the contributions of the features as follows: F3, F5, F8, F14, F10, F15, F4, F7, F9, F11, F6, F1, F2, F12, and F13. Among the top features, F3 is the only negative feature, increasing the probability of predicting the alternative label, C2. Other top features that are shifting the prediction towards C1 are F5, F8, and F14. Similar to F3, the features F4, F12, and F11 have negative contributions, supporting the generation of C2. By comparing the strong joint positive attribution to the joint negative attribution, it is evident why the classifier is very certain that C1 is the right label for this instance.

{'C1': '98.89%', 'C2': '1.11%'}

{'F3': 'Type of Travel', 'F5': 'Type Of Booking', 'F8': 'Common Room entertainment', 'F14': 'Stay comfort', 'F10': 'Cleanliness', 'F15': 'Hotel wifi service', 'F4': 'Other service', 'F7': 'Ease of Online booking', 'F9': 'Age', 'F11': 'Checkin\\/Checkout service', 'F6': 'Food and drink', 'F1': 'Departure\\/Arrival convenience', 'F2': 'purpose_of_travel', 'F12': 'Hotel location', 'F13': 'Gender'}

{'F3': 'F3', 'F4': 'F5', 'F12': 'F8', 'F11': 'F14', 'F15': 'F10', 'F6': 'F15', 'F14': 'F4', 'F8': 'F7', 'F5': 'F9', 'F13': 'F11', 'F10': 'F6', 'F7': 'F1', 'F2': 'F2', 'F9': 'F12', 'F1': 'F13'}

{'C2': 'C1', 'C1': 'C2'}

dissatisfied

{'C1': 'dissatisfied', 'C2': 'satisfied'}

RandomForestClassifier

C2

Used Cars Price-Range Prediction

The prediction probability associated with class C1 is 10.50%, while that of class C2 is 89.50%, therefore, it can be concluded that C2 is the most probable label for the given case according to the model. All the input features are shown to contribute to the above decision, and the ones with the strongest influence on the classification decision are F2, F7, and F8, but F1, F5, F9, and F3 are shown to be the least relevant features . Finally, the degree of influence of F4, F6, and F10 can be described as moderate. The model's high confidence can be attributed to the strong positive contributions of F7 and F2 which are supported by the contributions of the remaining positive features F4, F1, and F5. Conversely, shifting the prediction in favour of C1, the negative features F8, F6, F9, F10, and F3.

{'C1': '10.50%', 'C2': '89.50%'}

{'F7': 'Power', 'F2': 'car_age', 'F8': 'Transmission', 'F4': 'Fuel_Type', 'F6': 'Name', 'F10': 'Mileage', 'F1': 'Engine', 'F9': 'Owner_Type', 'F5': 'Kilometers_Driven', 'F3': 'Seats'}

{'F4': 'F7', 'F5': 'F2', 'F8': 'F8', 'F7': 'F4', 'F6': 'F6', 'F2': 'F10', 'F3': 'F1', 'F9': 'F9', 'F1': 'F5', 'F10': 'F3'}

{'C1': 'C1', 'C2': 'C2'}

High

{'C1': 'Low', 'C2': 'High'}

SVC

C1

Food Ordering Customer Churn Prediction

The model labels the case as C1 with fairly high confidence equal to 89.73%, whereas the likelihood of C2 is only 10.27%. Analysis shows that only 20 of the 46 input variables contribute to the prediction assertion above. The prediction judgement C1 is mainly based on the variables F38, F29, F30, and F46. F1, F11, F18, F15, F23, and F45 also contribute to the decision, however, their degree of influence is only moderate. According to the direction of influence analysis, F38, F46, F23, and F15 positively support the decision of the model to assign the label C1. However, F29, F11, F45, F30, F1, and F18 reduce the likelihood or chance that C1 is the true label for this particular test instance. The main variables with less influence on the above classification decision are F16, F26, F33, and F25.

{'C1': '89.73%', 'C2': '10.27%'}

{'F38': 'Ease and convenient', 'F29': 'Unaffordable', 'F46': 'Good Food quality', 'F30': 'Wrong order delivered', 'F1': 'Delay of delivery person picking up food', 'F11': 'Politeness', 'F18': 'Self Cooking', 'F45': 'Late Delivery', 'F15': 'Health Concern', 'F23': 'More Offers and Discount', 'F40': 'Easy Payment option', 'F39': 'Time saving', 'F14': 'Perference(P2)', 'F3': 'Gender', 'F10': 'Good Road Condition', 'F22': 'Google Maps Accuracy', 'F17': 'Good Taste ', 'F8': 'Good Tracking system', 'F42': 'Bad past experience', 'F31': 'Marital Status', 'F16': 'Influence of rating', 'F33': 'Delivery person ability', 'F26': 'Low quantity low time', 'F25': 'Age', 'F12': 'Less Delivery time', 'F34': 'High Quality of package', 'F28': 'Maximum wait time', 'F24': 'Number of calls', 'F36': 'Freshness ', 'F32': 'Temperature', 'F4': 'Residence in busy location', 'F41': 'Long delivery time', 'F6': 'Order Time', 'F13': 'Influence of time', 'F20': 'Order placed by mistake', 'F9': 'Missing item', 'F5': 'Delay of delivery person getting assigned', 'F43': 'Family size', 'F35': 'Unavailability', 'F2': 'Poor Hygiene', 'F37': 'More restaurant choices', 'F27': 'Perference(P1)', 'F7': 'Educational Qualifications', 'F44': 'Monthly Income', 'F21': 'Occupation', 'F19': 'Good Quantity'}

{'F10': 'F38', 'F23': 'F29', 'F15': 'F46', 'F27': 'F30', 'F26': 'F1', 'F42': 'F11', 'F17': 'F18', 'F19': 'F45', 'F18': 'F15', 'F14': 'F23', 'F13': 'F40', 'F11': 'F39', 'F9': 'F14', 'F2': 'F3', 'F35': 'F10', 'F34': 'F22', 'F45': 'F17', 'F16': 'F8', 'F21': 'F42', 'F3': 'F31', 'F38': 'F16', 'F37': 'F33', 'F36': 'F26', 'F1': 'F25', 'F39': 'F12', 'F40': 'F34', 'F32': 'F28', 'F41': 'F24', 'F43': 'F36', 'F44': 'F32', 'F33': 'F4', 'F24': 'F41', 'F31': 'F6', 'F30': 'F13', 'F29': 'F20', 'F28': 'F9', 'F25': 'F5', 'F7': 'F43', 'F22': 'F35', 'F20': 'F2', 'F12': 'F37', 'F8': 'F27', 'F6': 'F7', 'F5': 'F44', 'F4': 'F21', 'F46': 'F19'}

{'C2': 'C1', 'C1': 'C2'}

Return

{'C1': 'Return', 'C2': 'Go Away'}

MLPClassifier

C2

Annual Income Earnings

The label predicted for this case is C2 with very high confidence of approximately 97.71% which insinuates that there is a marginal possibility that C1 could be the label. The above classification decision is largely due to the values of F9, F11, F4, and F12. On the other hand, F6 and F3 are less relevant when the model is deciding the correct label for the case here. Digging deeper revealed that each feature either positively or negatively contribute to the prediction made here. Six features contradicted the classification decision, while the remaining ones positively supported the C2 prediction. The negative features driving the prediction towards C1 are F4, F7, F5, F8, F13, and F3 and countering their influence are the top positive features are F9, F11, F10, and F12.

{'C1': '2.29%', 'C2': '97.71%'}

{'F9': 'Capital Gain', 'F11': 'Marital Status', 'F4': 'Capital Loss', 'F12': 'Relationship', 'F7': 'Hours per week', 'F10': 'Education', 'F2': 'Country', 'F14': 'Age', 'F1': 'Occupation', 'F5': 'Sex', 'F8': 'Education-Num', 'F13': 'Workclass', 'F6': 'fnlwgt', 'F3': 'Race'}

{'F11': 'F9', 'F6': 'F11', 'F12': 'F4', 'F8': 'F12', 'F13': 'F7', 'F4': 'F10', 'F14': 'F2', 'F1': 'F14', 'F7': 'F1', 'F10': 'F5', 'F5': 'F8', 'F2': 'F13', 'F3': 'F6', 'F9': 'F3'}

{'C2': 'C1', 'C1': 'C2'}

Above 50K

{'C1': 'Under 50K', 'C2': 'Above 50K'}

SVM_linear

C1

Wine Quality Prediction

The likelihood of C1 being the correct label for the selected case or instance is 67.54% according to the classifier. This means, there is a 32.46% chance that C2 could be the label and the classification assertion above is influenced mainly by the variables F11, F7, F1, and F10. On the contrary, F4, F8, and F6 are deemed less important when deciding the correct label for this given case. Decreasing the likelihood of the predicted label , C1, are the variables F10, F5, F8, and F6, therefore, these negative variables support the alternative class C2. However, the collective or joint attribution of the top positive variables, F7, F11, and F1 is strong enough to tilt the classification in favour of C1.

{'C2': '32.46%', 'C1': '67.54%'}

{'F11': 'residual sugar', 'F7': 'volatile acidity', 'F1': 'alcohol', 'F10': 'fixed acidity', 'F2': 'chlorides', 'F5': 'sulphates', 'F9': 'citric acid', 'F3': 'free sulfur dioxide', 'F4': 'density', 'F8': 'total sulfur dioxide', 'F6': 'pH'}

{'F4': 'F11', 'F2': 'F7', 'F11': 'F1', 'F1': 'F10', 'F5': 'F2', 'F10': 'F5', 'F3': 'F9', 'F6': 'F3', 'F8': 'F4', 'F7': 'F8', 'F9': 'F6'}

{'C1': 'C2', 'C2': 'C1'}

high quality

{'C2': 'low_quality', 'C1': 'high quality'}

KNeighborsClassifier

C1

Credit Risk Classification

The confidence level score with respect to each class label suggests that this case should be labelled as C1. Specifically, there is about an 80.0% chance that C1 is the correct label. However, this implies that there is also about a 20.0% chance that it should be C2. The above prediction decision is based predominantly on the influence of the following features: F2, F5, F7, F1, F10, F4, and F9. According to the analysis, the features F2, F5, and F7 have a very strong positive influence, swinging the prediction decision towards C1. In contrast, the value of F1 also suggests the decision should be the alternative class, C2. Similar to F1, the values of F11, F10, and F4 indicate the label could be C2. However, the influence of these features is very small compared to F2, F5, F7, and F1. Finally, the attributes with a moderately low influence on the final prediction decision for this case include F9, F8, F6, and F3. The values of F9 and F3 have a negative attribution, while F8 and F6 have positive attributions.

{'C1': '80.00%', 'C2': '20.00%'}

{'F2': 'fea_4', 'F5': 'fea_8', 'F7': 'fea_2', 'F1': 'fea_9', 'F11': 'fea_6', 'F10': 'fea_10', 'F4': 'fea_1', 'F9': 'fea_11', 'F8': 'fea_7', 'F3': 'fea_3', 'F6': 'fea_5'}

{'F4': 'F2', 'F8': 'F5', 'F2': 'F7', 'F9': 'F1', 'F6': 'F11', 'F10': 'F10', 'F1': 'F4', 'F11': 'F9', 'F7': 'F8', 'F3': 'F3', 'F5': 'F6'}

{'C2': 'C1', 'C1': 'C2'}

Low

{'C1': 'Low', 'C2': 'High'}

BernoulliNB

C2

Job Change of Data Scientists

The classification algorithm is pretty confident that the correct label for the data under consideration is C2, wowever, it is noteworthy to consider that C1 has about a 15.13% chance of being the correct label. The predicted probability of each label is assigned based on the influence of features such as F8, F5, F11, and F4. However, the analysis shows that the values of F12, F7, F2, and F3 are less relevant when classifying the data. Only the features F11, F10, F12, F7, F2, and F3 have negative attributions, decreasing the predicted probability of the assigned label and one can say these features are shifting the prediction decision towards the label C1.

{'C1': '15.13%', 'C2': '84.87%'}

{'F8': 'city', 'F5': 'enrolled_university', 'F11': 'relevent_experience', 'F4': 'city_development_index', 'F10': 'experience', 'F9': 'education_level', 'F1': 'major_discipline', 'F6': 'last_new_job', 'F12': 'gender', 'F7': 'company_size', 'F2': 'company_type', 'F3': 'training_hours'}

{'F3': 'F8', 'F6': 'F5', 'F5': 'F11', 'F1': 'F4', 'F9': 'F10', 'F7': 'F9', 'F8': 'F1', 'F12': 'F6', 'F4': 'F12', 'F10': 'F7', 'F11': 'F2', 'F2': 'F3'}

{'C2': 'C1', 'C1': 'C2'}

Leave

{'C1': 'Stay', 'C2': 'Leave'}

GradientBoostingClassifier

C2

Printer Sales

The case, despite having features with considerable negative impact, also has numerous and measurable positive features, so the assignment of the label C2 by the model is very likely since the predicted probability is 91.95% which is very higher than that of C1. The F20, F13, and F7 were the most important features driving the model to arrive at the labelling assignment of C2. F11 and F23 have nearly identical positive attributions, while F26 and F22 has negative impacts, swinging the prediction towards a different label. However, the joint positive impact of F11, F20, F7, and F23 stands out over the impact of F13, F24, F22, and F26, favouring the prediction of the C2 model. All things considered, there are more features with a positive impact than those with negative impact; the mean attribution of the positive attributes is much larger which somewhat explains why the confidence level is very high. Above all, it is important to note that the prediction is made with less emphasis on the values of F3, F21, F4, and F12 hence they are practically irrelevant when it comes to labelling this case.

{'C1': '8.05%', 'C2': '91.95%'}

{'F20': 'X24', 'F13': 'X8', 'F7': 'X1', 'F24': 'X21', 'F22': 'X4', 'F23': 'X6', 'F11': 'X3', 'F26': 'X22', 'F2': 'X7', 'F10': 'X15', 'F1': 'X20', 'F5': 'X11', 'F25': 'X10', 'F8': 'X19', 'F14': 'X5', 'F18': 'X16', 'F15': 'X23', 'F6': 'X9', 'F19': 'X17', 'F9': 'X18', 'F21': 'X25', 'F16': 'X14', 'F12': 'X2', 'F4': 'X13', 'F17': 'X12', 'F3': 'X26'}

{'F24': 'F20', 'F8': 'F13', 'F1': 'F7', 'F21': 'F24', 'F4': 'F22', 'F6': 'F23', 'F3': 'F11', 'F22': 'F26', 'F7': 'F2', 'F15': 'F10', 'F20': 'F1', 'F11': 'F5', 'F10': 'F25', 'F19': 'F8', 'F5': 'F14', 'F16': 'F18', 'F23': 'F15', 'F9': 'F6', 'F17': 'F19', 'F18': 'F9', 'F25': 'F21', 'F14': 'F16', 'F2': 'F12', 'F13': 'F4', 'F12': 'F17', 'F26': 'F3'}

{'C1': 'C1', 'C2': 'C2'}

More

{'C1': 'Less', 'C2': 'More'}

LogisticRegression

C1

Cab Surge Pricing System

The predicted label is C1 given the predictability of C2 is 28.96% and that of C3 is 23.41%. Considering the probabilities of the classes, the model can be described as being moderately confident. The prediction of C1 can be attributed to the varying degree of contributions of the input features. Attribution analysis indicates that F4, F11, and F8 are considered the most influential. Those with moderate influence are F9, F1, F5, F10, F7, and F2, whereas on the contrary, the least influential ones are F6, F12, and F3. The analysis also revealed that not all the features contribute positively to the prediction decision and amongst the input features, the ones with negative attributions decreasing the likelihood of the C1 prediction are F11, F8, F5, F10, and F7 whereas conversely, the top positive features are F4, F9, and F1.

{'C2': '28.96%', 'C3': '23.41%', 'C1': '47.63%'}

{'F4': 'Type_of_Cab', 'F11': 'Confidence_Life_Style_Index', 'F8': 'Destination_Type', 'F9': 'Trip_Distance', 'F1': 'Cancellation_Last_1Month', 'F5': 'Life_Style_Index', 'F10': 'Customer_Rating', 'F7': 'Var3', 'F2': 'Var1', 'F6': 'Customer_Since_Months', 'F12': 'Var2', 'F3': 'Gender'}

{'F2': 'F4', 'F5': 'F11', 'F6': 'F8', 'F1': 'F9', 'F8': 'F1', 'F4': 'F5', 'F7': 'F10', 'F11': 'F7', 'F9': 'F2', 'F3': 'F6', 'F10': 'F12', 'F12': 'F3'}

{'C3': 'C2', 'C1': 'C3', 'C2': 'C1'}

C3

{'C2': 'Low', 'C3': 'Medium', 'C1': 'High'}

RandomForestClassifier

C1

Company Bankruptcy Prediction

The model outputs a predicted probability of 2.55% for the C2 label and 97.45% for the C1 label. Judging from above, the most probable class is C1. Hence, C1 is the assigned label by the model, with a very high confidence level. The top features contributing to the prediction assessment above are F42, F13, F37, F20, and F53. However, the values of about twenty features are deemed relevant while the remaining are regarded as irrelevant when classifying the given case. These irrelevant features include F65, F5, F39, and F85. Among the relevant features, F37, F60, F7, F46, F78, and F19 are shown to be the only positive features that increase the model's response in favour of the assigned label C1. In contrast, the majority of the relevant features, mainly F42, F13, F20, and F53, have negative contributions, decreasing the odds of the label C1, hence supporting the assignment of C2 to the given case.

{'C2': '2.55%', 'C1': '97.45%'}

{'F42': " Net Income to Stockholder's Equity", 'F13': ' Total income\\/Total expense', 'F37': ' Borrowing dependency', 'F20': ' Continuous interest rate (after tax)', 'F53': ' Net Value Per Share (B)', 'F32': ' Cash\\/Current Liability', 'F48': ' Net worth\\/Assets', 'F60': ' Fixed Assets Turnover Frequency', 'F7': ' Interest-bearing debt interest rate', 'F46': ' No-credit Interval', 'F51': ' Net Value Per Share (A)', 'F79': ' Long-term fund suitability ratio (A)', 'F78': ' Equity to Long-term Liability', 'F75': ' Realized Sales Gross Margin', 'F19': ' Current Asset Turnover Rate', 'F47': ' Working Capital to Total Assets', 'F27': ' Interest Coverage Ratio (Interest expense to EBIT)', 'F66': ' Working capitcal Turnover Rate', 'F74': ' Inventory Turnover Rate (times)', 'F45': ' After-tax net Interest Rate', 'F65': ' Working Capital\\/Equity', 'F5': ' Liability to Equity', 'F39': ' Operating Gross Margin', 'F85': ' Cash Flow Per Share', 'F87': ' Contingent liabilities\\/Net worth', 'F29': ' Operating Profit Per Share (Yuan ¥)', 'F50': ' Operating Profit Rate', 'F40': ' Net Worth Turnover Rate (times)', 'F61': ' Continuous Net Profit Growth Rate', 'F83': ' Long-term Liability to Current Assets', 'F33': ' Fixed Assets to Assets', 'F23': ' Inventory and accounts receivable\\/Net value', 'F22': ' Regular Net Profit Growth Rate', 'F54': ' Current Liability to Equity', 'F44': ' Equity to Liability', 'F73': ' Current Liability to Liability', 'F80': ' Operating profit\\/Paid-in capital', 'F24': ' Net Value Per Share (C)', 'F58': ' Operating Funds to Liability', 'F9': ' Current Liability to Current Assets', 'F17': ' Current Ratio', 'F92': ' Quick Assets\\/Current Liability', 'F64': ' Tax rate (A)', 'F56': ' After-tax Net Profit Growth Rate', 'F15': ' Per Share Net profit before tax (Yuan ¥)', 'F4': ' Total Asset Turnover', 'F1': ' CFO to Assets', 'F16': ' Cash Reinvestment %', 'F28': ' Net profit before tax\\/Paid-in capital', 'F38': ' Cash Flow to Equity', 'F52': ' Debt ratio %', 'F26': ' Current Liabilities\\/Liability', 'F84': ' Interest Expense Ratio', 'F91': ' Cash Flow to Sales', 'F36': ' Total Asset Growth Rate', 'F2': ' Inventory\\/Current Liability', 'F86': ' Allocation rate per person', 'F18': ' Operating Expense Rate', 'F31': ' Operating profit per person', 'F34': ' Net Income to Total Assets', 'F72': ' Net Value Growth Rate', 'F21': ' ROA(B) before interest and depreciation after tax', 'F69': ' Cash Flow to Liability', 'F41': ' Inventory\\/Working Capital', 'F3': ' Retained Earnings to Total Assets', 'F77': ' Total assets to GNP price', 'F57': ' Persistent EPS in the Last Four Seasons', 'F6': ' Total debt\\/Total net worth', 'F30': ' Quick Ratio', 'F93': ' Revenue per person', 'F88': ' Non-industry income and expenditure\\/revenue', 'F25': ' Cash\\/Total Assets', 'F8': ' ROA(A) before interest and % after tax', 'F62': ' ROA(C) before interest and depreciation before interest', 'F12': ' Research and development expense rate', 'F82': ' Cash Flow to Total Assets', 'F49': ' Pre-tax net Interest Rate', 'F90': ' Accounts Receivable Turnover', 'F76': ' Current Liability to Assets', 'F59': ' Quick Assets\\/Total Assets', 'F11': ' Total expense\\/Assets', 'F10': ' Operating Profit Growth Rate', 'F67': ' Average Collection Days', 'F35': ' Current Assets\\/Total Assets', 'F89': ' Current Liabilities\\/Equity', 'F81': ' Realized Sales Gross Profit Growth Rate', 'F14': ' Cash flow rate', 'F71': ' Total Asset Return Growth Rate Ratio', 'F55': ' Degree of Financial Leverage (DFL)', 'F43': ' Cash Turnover Rate', 'F68': ' Quick Asset Turnover Rate', 'F70': ' Revenue Per Share (Yuan ¥)', 'F63': ' Gross Profit to Sales'}

{'F59': 'F42', 'F57': 'F13', 'F3': 'F37', 'F12': 'F20', 'F27': 'F53', 'F32': 'F32', 'F84': 'F48', 'F22': 'F60', 'F1': 'F7', 'F56': 'F46', 'F42': 'F51', 'F52': 'F79', 'F23': 'F78', 'F83': 'F75', 'F61': 'F19', 'F67': 'F47', 'F60': 'F27', 'F73': 'F66', 'F18': 'F74', 'F79': 'F45', 'F68': 'F65', 'F66': 'F5', 'F62': 'F39', 'F65': 'F85', 'F64': 'F87', 'F63': 'F29', 'F58': 'F50', 'F55': 'F40', 'F54': 'F61', 'F69': 'F83', 'F74': 'F33', 'F70': 'F23', 'F85': 'F22', 'F92': 'F54', 'F91': 'F44', 'F90': 'F73', 'F89': 'F80', 'F88': 'F24', 'F87': 'F58', 'F86': 'F9', 'F82': 'F17', 'F71': 'F92', 'F81': 'F64', 'F80': 'F56', 'F78': 'F15', 'F77': 'F4', 'F76': 'F1', 'F75': 'F16', 'F72': 'F28', 'F53': 'F38', 'F47': 'F52', 'F51': 'F26', 'F14': 'F84', 'F25': 'F91', 'F24': 'F36', 'F21': 'F2', 'F20': 'F86', 'F19': 'F18', 'F17': 'F31', 'F16': 'F34', 'F15': 'F72', 'F13': 'F21', 'F50': 'F69', 'F11': 'F41', 'F10': 'F3', 'F9': 'F77', 'F8': 'F57', 'F7': 'F6', 'F6': 'F30', 'F5': 'F93', 'F4': 'F88', 'F26': 'F25', 'F28': 'F8', 'F29': 'F62', 'F30': 'F12', 'F49': 'F82', 'F48': 'F49', 'F2': 'F90', 'F46': 'F76', 'F45': 'F59', 'F44': 'F11', 'F43': 'F10', 'F41': 'F67', 'F40': 'F35', 'F39': 'F89', 'F38': 'F81', 'F37': 'F14', 'F36': 'F71', 'F35': 'F55', 'F34': 'F43', 'F33': 'F68', 'F31': 'F70', 'F93': 'F63'}

{'C1': 'C2', 'C2': 'C1'}

Yes

{'C2': 'No', 'C1': 'Yes'}

SVC

C1

Broadband Sevice Signup

The algorithm identifies the provided data or case as C1 with a greater level of certainty since the prediction probability of class C2 is just 0.07 percent as a result, C2 is less likely than C1. The influence of input features such as F21, F32, F18, F9, and F17 is mostly responsible for the classification verdict above with only F17 having a negative influence among them, slightly pulling the decision in favour of C2. F21, F32, F18, and F9, on the other hand, make considerable positive contributions in favour of assigning C1 to the data. F34, F37, F40, F23, F24, F11, F35, and F39 are some more features that have a modest effect on the algorithm's decision. But, not all features are demonstrated to influence the classification decision either negatively or positively to the aforementioned classification outcome and in reality, a number of these are demonstrated to be irrelevant for determining the suitable label for this case and these include F7, F16, F13, and F28. All in all, the most important features for this classification instance are F21 and F32, whereas F5 and F36 are the least important.

{'C1': '99.93%', 'C2': '0.07%'}

{'F21': 'X38', 'F32': 'X32', 'F18': 'X31', 'F9': 'X25', 'F17': 'X8', 'F34': 'X35', 'F37': 'X1', 'F35': 'X3', 'F23': 'X28', 'F39': 'X19', 'F24': 'X9', 'F40': 'X11', 'F11': 'X10', 'F31': 'X21', 'F27': 'X17', 'F38': 'X4', 'F30': 'X36', 'F3': 'X2', 'F4': 'X6', 'F6': 'X34', 'F16': 'X37', 'F7': 'X40', 'F13': 'X42', 'F28': 'X41', 'F42': 'X5', 'F15': 'X33', 'F26': 'X39', 'F29': 'X24', 'F25': 'X30', 'F19': 'X27', 'F2': 'X26', 'F8': 'X23', 'F10': 'X22', 'F14': 'X20', 'F22': 'X18', 'F12': 'X16', 'F1': 'X15', 'F33': 'X14', 'F36': 'X13', 'F5': 'X12', 'F41': 'X7', 'F20': 'X29'}

{'F35': 'F21', 'F29': 'F32', 'F28': 'F18', 'F23': 'F9', 'F6': 'F17', 'F32': 'F34', 'F40': 'F37', 'F2': 'F35', 'F26': 'F23', 'F17': 'F39', 'F7': 'F24', 'F9': 'F40', 'F8': 'F11', 'F19': 'F31', 'F15': 'F27', 'F3': 'F38', 'F33': 'F30', 'F1': 'F3', 'F4': 'F4', 'F31': 'F6', 'F34': 'F16', 'F37': 'F7', 'F38': 'F13', 'F39': 'F28', 'F41': 'F42', 'F30': 'F15', 'F36': 'F26', 'F22': 'F29', 'F27': 'F25', 'F25': 'F19', 'F24': 'F2', 'F21': 'F8', 'F20': 'F10', 'F18': 'F14', 'F16': 'F22', 'F14': 'F12', 'F13': 'F1', 'F12': 'F33', 'F11': 'F36', 'F10': 'F5', 'F5': 'F41', 'F42': 'F20'}

{'C1': 'C1', 'C2': 'C2'}

No

{'C1': 'No', 'C2': 'Yes'}

SVM

C2

Customer Churn Modelling

For the given dataset instance, the label assigned by the classifier is C2 since it has a predicted probability of about 89.16%. On the other hand, there is a 9.0% chance that C1 could be the appropriate label, whereas C3 only has a 1.84% chance of being the true label. The classifier arrived at this classification verdict chiefly due to the influence and contributions of variables such as F7, F1, F3, and F10. However, there is less emphasis on the values of F2, F4, and F5, since their impact on the classifier with respect to the given case is smaller compared to the other variables, hence they are the least ranked features. From the attribution analysis, there are four variables with negative contributions, pushing the verdict in the direction of C1. These negative variables are F7, F10, F9, and F8, and their influence on the classifier could explain why there is a little bit of doubt about the correctness of the C2 class assigned and the notable positive variables are F1, F6, F2, and F3.

{'C2': '89.16%', 'C1': '9.0%', 'C3': '1.84%'}

{'F7': 'IsActiveMember', 'F1': 'Age', 'F3': 'Geography', 'F10': 'NumOfProducts', 'F9': 'Gender', 'F6': 'Tenure', 'F8': 'CreditScore', 'F2': 'Balance', 'F4': 'EstimatedSalary', 'F5': 'HasCrCard'}

{'F9': 'F7', 'F4': 'F1', 'F2': 'F3', 'F7': 'F10', 'F3': 'F9', 'F5': 'F6', 'F1': 'F8', 'F6': 'F2', 'F10': 'F4', 'F8': 'F5'}

{'C1': 'C2', 'C2': 'C1', 'C3': 'C3'}

Stay

{'C2': 'Stay', 'C1': 'Leave', 'C3': 'Other'}

GradientBoostingClassifier

C2

Basketball Players Career Length Prediction

The case is labelled as C2 by the model but looking at the predicted probabilities across the different classes, there is a 33.63% chance that the label could be C1. To explain the above prediction conclusion, the analysis revealed that the majority of the features have negative influences or attributions, pushing the prediction away from C2 in favour of C1. The negative features include F5, F4, F14, F19, and F16 and the values of these features are ranked higher than any of the positive features. Shifting the prediction in the direction of C2 are the positive features F10, F8, F2, and F1. The analysis also revealed that the values of F9, F11, and F18 are less relevant to the prediction for the case under consideration.

{'C1': '33.63%', 'C2': '66.37%'}

{'F5': 'GamesPlayed', 'F4': 'OffensiveRebounds', 'F14': 'FieldGoalPercent', 'F19': 'FreeThrowPercent', 'F16': '3PointPercent', 'F10': '3PointAttempt', 'F6': 'FieldGoalsMade', 'F7': 'Blocks', 'F12': 'DefensiveRebounds', 'F15': 'Turnovers', 'F8': 'Rebounds', 'F3': 'FreeThrowAttempt', 'F2': 'MinutesPlayed', 'F13': 'Assists', 'F17': 'FieldGoalsAttempt', 'F1': '3PointMade', 'F9': 'PointsPerGame', 'F11': 'FreeThrowMade', 'F18': 'Steals'}

{'F1': 'F5', 'F13': 'F4', 'F6': 'F14', 'F12': 'F19', 'F9': 'F16', 'F8': 'F10', 'F4': 'F6', 'F18': 'F7', 'F14': 'F12', 'F19': 'F15', 'F15': 'F8', 'F11': 'F3', 'F2': 'F2', 'F16': 'F13', 'F5': 'F17', 'F7': 'F1', 'F3': 'F9', 'F10': 'F11', 'F17': 'F18'}

{'C1': 'C1', 'C2': 'C2'}

Less than 5

{'C1': 'More than 5', 'C2': 'Less than 5'}

BernoulliNB

C1

Customer Churn Modelling

The most likely label chosen by the model in this case is C1. The decision above is based on the prediction probabilities for the two possible labels, C1 and C2, which are 94.25% and 5.75%, respectively. The following variables can be ranked from most important to least important based on their contribution to the model when it comes to this instance: F8, F4, F7, F3, F1, F10, F6, F5, F2, and F9. F4 and F8 turned out to be the most important positive variables, supporting the model towards assigning the class C1. The least positive variables are F5 and F6, which have less effect on the model. In fact, most of the input features have negative contributions towards the assignment of class C1, leading to a decision change in favour of the other label, C2. The most negative variables are F1, F7, and F3, and the least negative are F2 and F9.

{'C1': '94.25%', 'C2': '5.75%'}

{'F8': 'IsActiveMember', 'F4': 'NumOfProducts', 'F7': 'Gender', 'F3': 'Geography', 'F1': 'Age', 'F10': 'CreditScore', 'F6': 'EstimatedSalary', 'F5': 'Balance', 'F2': 'HasCrCard', 'F9': 'Tenure'}

{'F9': 'F8', 'F7': 'F4', 'F3': 'F7', 'F2': 'F3', 'F4': 'F1', 'F1': 'F10', 'F10': 'F6', 'F6': 'F5', 'F8': 'F2', 'F5': 'F9'}

{'C1': 'C1', 'C2': 'C2'}

Stay

{'C1': 'Stay', 'C2': 'Leave'}

BernoulliNB

C2

Personal Loan Modelling

As per the classification algorithm employed, the most probable label for the data under consideration is C2 since the chances of C1 is very slim and negligible. The main driver behind the labelling decision above is F1. The features with moderate influence are F3, F7, F8, F4, F5, and F9, while those with very small or marginal impact are F6 and F2. The direction of influence of the input features could be used to explain why the algorithm is very confident here. Most of the features have a positive impact, increasing or improving the chances of C2 being the correct label and the feature with a significantly higher contribution, F1, is a positive feature which when coupled with other positives F7, F8, F9, and F5 encourages the prediction or assignment of the C2 label. Furthermore, aside from F3 and F4, the other two negative features, F6 and F2, are shown to have a significantly lower impact on the algorithm and the very marginal doubt in the decision can be attributed to the influence of the negative features.

{'C2': '99.99%', 'C1': '0.01%'}

{'F1': 'CD Account', 'F3': 'Income', 'F7': 'CCAvg', 'F8': 'Securities Account', 'F4': 'Education', 'F5': 'Family', 'F9': 'Mortgage', 'F6': 'Age', 'F2': 'Extra_service'}

{'F8': 'F1', 'F2': 'F3', 'F4': 'F7', 'F7': 'F8', 'F5': 'F4', 'F3': 'F5', 'F6': 'F9', 'F1': 'F6', 'F9': 'F2'}

{'C2': 'C2', 'C1': 'C1'}

Reject

{'C2': 'Reject', 'C1': 'Accept'}

MLPClassifier

C1

Vehicle Insurance Claims

The ML algorithm classifies the provided data or case as C1 with a likelihood of 80.70%, hinting that the likelihood of C2 being the correct label is only 19.30%. This classification decision above is mainly based on the influence or contributions of the input features. The most relevant features driving the classification algorithm to arrive at the above decision are F17, F3, F10, F11, F28, F15, and F13. On the other side, not all of the input features are considered relevant when deciding the appropriate label for the given data instance, and these irrelevant features include F9, F4, F19, F21, and F27. Among the top influential features, F28, F15, and F13 are regarded as negative features since their contributions push the algorithm's decision towards the less likely class, C2, although F17, F3, F10, F31, and F11 have positive contributions, increasing the probability that C1 is the right label here.

{'C2': '19.30%', 'C1': '80.70%'}

{'F17': 'incident_severity', 'F3': 'insured_relationship', 'F11': 'authorities_contacted', 'F10': 'vehicle_claim', 'F15': 'umbrella_limit', 'F13': 'insured_hobbies', 'F28': 'incident_type', 'F31': 'policy_deductable', 'F23': 'auto_make', 'F5': 'number_of_vehicles_involved', 'F29': 'insured_occupation', 'F30': 'property_damage', 'F2': 'incident_state', 'F18': 'auto_year', 'F16': 'capital-loss', 'F20': 'policy_csl', 'F33': 'collision_type', 'F14': 'capital-gains', 'F32': 'property_claim', 'F7': 'incident_hour_of_the_day', 'F9': 'police_report_available', 'F4': 'policy_annual_premium', 'F19': 'incident_city', 'F27': 'insured_zip', 'F21': 'bodily_injuries', 'F22': 'injury_claim', 'F1': 'witnesses', 'F26': 'total_claim_amount', 'F25': 'insured_education_level', 'F8': 'insured_sex', 'F24': 'policy_state', 'F6': 'age', 'F12': 'months_as_customer'}

{'F27': 'F17', 'F24': 'F3', 'F28': 'F11', 'F16': 'F10', 'F5': 'F15', 'F23': 'F13', 'F25': 'F28', 'F3': 'F31', 'F33': 'F23', 'F10': 'F5', 'F22': 'F29', 'F31': 'F30', 'F29': 'F2', 'F17': 'F18', 'F8': 'F16', 'F19': 'F20', 'F26': 'F33', 'F7': 'F14', 'F15': 'F32', 'F9': 'F7', 'F32': 'F9', 'F4': 'F4', 'F30': 'F19', 'F6': 'F27', 'F11': 'F21', 'F14': 'F22', 'F12': 'F1', 'F13': 'F26', 'F21': 'F25', 'F20': 'F8', 'F18': 'F24', 'F2': 'F6', 'F1': 'F12'}

{'C1': 'C2', 'C2': 'C1'}

Fraud

{'C2': 'Not Fraud', 'C1': 'Fraud'}

SVC

C1

Broadband Sevice Signup

The predicted probability of class C2 is 12.81% and that of class C1 is 87.19%. Therefore, the label chosen by the model is C1, which is the most probable class. The top two features with significant influence on the prediction verdict above are F2 and F40. These features have positive attributions, shifting the decision higher in support of label C1. Other positive features are F22, F29, F5, and F13. Decreasing the likelihood of the assigned label are the negative features such as F11, F36, F20, and F7. Finally, the values of features such as F30, F41, F39, F14, F42, and F27 are considered irrelevant to the prediction decision above.

{'C2': '12.81%', 'C1': '87.19%'}

{'F2': 'X38', 'F40': 'X32', 'F11': 'X22', 'F22': 'X35', 'F29': 'X25', 'F36': 'X16', 'F20': 'X12', 'F7': 'X31', 'F10': 'X3', 'F8': 'X9', 'F3': 'X1', 'F15': 'X19', 'F16': 'X4', 'F5': 'X2', 'F31': 'X29', 'F28': 'X42', 'F13': 'X36', 'F12': 'X21', 'F23': 'X40', 'F1': 'X10', 'F30': 'X33', 'F41': 'X5', 'F39': 'X6', 'F14': 'X41', 'F27': 'X39', 'F42': 'X7', 'F33': 'X37', 'F26': 'X8', 'F34': 'X34', 'F19': 'X18', 'F21': 'X17', 'F32': 'X11', 'F4': 'X30', 'F6': 'X28', 'F25': 'X27', 'F38': 'X26', 'F18': 'X13', 'F17': 'X14', 'F9': 'X23', 'F37': 'X15', 'F35': 'X20', 'F24': 'X24'}

{'F35': 'F2', 'F29': 'F40', 'F20': 'F11', 'F32': 'F22', 'F23': 'F29', 'F14': 'F36', 'F10': 'F20', 'F28': 'F7', 'F2': 'F10', 'F7': 'F8', 'F40': 'F3', 'F17': 'F15', 'F3': 'F16', 'F1': 'F5', 'F42': 'F31', 'F38': 'F28', 'F33': 'F13', 'F19': 'F12', 'F37': 'F23', 'F8': 'F1', 'F30': 'F30', 'F41': 'F41', 'F4': 'F39', 'F39': 'F14', 'F36': 'F27', 'F5': 'F42', 'F34': 'F33', 'F6': 'F26', 'F31': 'F34', 'F16': 'F19', 'F15': 'F21', 'F9': 'F32', 'F27': 'F4', 'F26': 'F6', 'F25': 'F25', 'F24': 'F38', 'F11': 'F18', 'F12': 'F17', 'F21': 'F9', 'F13': 'F37', 'F18': 'F35', 'F22': 'F24'}

{'C2': 'C2', 'C1': 'C1'}

Yes

{'C2': 'No', 'C1': 'Yes'}

MLPClassifier

C2

Ethereum Fraud Detection

The C1 has a predicted probability of just 3.10% while that of the C2 is 96.90%, therefore, the most likely class selected by the classifier for the given data is C2. The relevant features contributing to this classification are mainly F25, F19, F16, F31, F7, F15, F12, F6, F17, F22, F18, F24, F8, F33, F20, F14, F32, F28, F13, and F5. As per the attribution analysis, F25 and F19 have a very strong joint positive contribution, increasing the classifier's response higher in favour of C2 than C1. In contrast, F16, F7, and F31 are the top negative features, degrading the classifier's response in favour of C1. Comparing the attributions of F25, F15, and F19 to those of the negative features mentioned above, it is not surprising that the classifier is quite confident that C2 is the most probable label here.

{'C1': '3.10%', 'C2': '96.90%'}

{'F25': 'Unique Received From Addresses', 'F19': ' ERC20 total Ether sent contract', 'F16': 'total ether received', 'F31': 'Sent tnx', 'F7': 'Number of Created Contracts', 'F15': ' ERC20 uniq rec token name', 'F12': ' ERC20 uniq rec contract addr', 'F6': 'max value received ', 'F17': 'total transactions (including tnx to create contract', 'F22': ' ERC20 uniq sent addr.1', 'F18': ' ERC20 uniq sent addr', 'F24': 'Received Tnx', 'F8': 'avg val received', 'F33': ' ERC20 uniq rec addr', 'F20': 'avg val sent', 'F14': 'min value received', 'F32': 'Unique Sent To Addresses', 'F28': ' ERC20 uniq sent token name', 'F13': 'Avg min between received tnx', 'F5': 'Time Diff between first and last (Mins)', 'F30': ' ERC20 min val rec', 'F2': ' ERC20 max val rec', 'F3': ' ERC20 min val sent', 'F4': ' ERC20 max val sent', 'F21': ' ERC20 avg val sent', 'F37': ' ERC20 avg val rec', 'F9': ' Total ERC20 tnxs', 'F34': ' ERC20 total ether sent', 'F38': ' ERC20 total Ether received', 'F29': 'total ether balance', 'F26': 'total ether sent contracts', 'F35': 'total Ether sent', 'F10': 'avg value sent to contract', 'F36': 'max val sent to contract', 'F11': 'min value sent to contract', 'F23': 'max val sent', 'F27': 'min val sent', 'F1': 'Avg min between sent tnx'}

{'F7': 'F25', 'F26': 'F19', 'F20': 'F16', 'F4': 'F31', 'F6': 'F7', 'F38': 'F15', 'F30': 'F12', 'F10': 'F6', 'F18': 'F17', 'F29': 'F22', 'F27': 'F18', 'F5': 'F24', 'F11': 'F8', 'F28': 'F33', 'F14': 'F20', 'F9': 'F14', 'F8': 'F32', 'F37': 'F28', 'F2': 'F13', 'F3': 'F5', 'F31': 'F30', 'F32': 'F2', 'F34': 'F3', 'F35': 'F4', 'F36': 'F21', 'F33': 'F37', 'F23': 'F9', 'F25': 'F34', 'F24': 'F38', 'F22': 'F29', 'F21': 'F26', 'F19': 'F35', 'F17': 'F10', 'F16': 'F36', 'F15': 'F11', 'F13': 'F23', 'F12': 'F27', 'F1': 'F1'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

RandomForestClassifier

C2

Printer Sales

There is only a 17.0% chance that C1 is the correct label which implies that the most probable label for the given data or case is C2 given its predicted likelihood of 83.0%. The main influential features resulting in the classification conclusions above are F3, F1, and F5 whereas the remaining features have either a moderate or negligible influence on the classifier. When it comes to assigning a label to this case, the classifier likely ignored the values of F20, F16, F2, F24, F22, and F17 since their respective degrees of influence are very close to zero. Among the influential features, only F23, F26, F13, F15, F21, and F9 are considered negative features mainly due to the fact that their contributions towards the decision here only serve to decrease the likelihood that C2 is the correct label and it can be said that these features favour labelling the case as C1. The remaining features such as F3, F1, F5, F4, F7, F11, and F12, offer positive contributions, increasing the likelihood of the C2 class.

{'C2': '83.00%', 'C1': '17.00%'}

{'F3': 'X8', 'F1': 'X24', 'F5': 'X1', 'F4': 'X2', 'F7': 'X10', 'F11': 'X15', 'F23': 'X25', 'F12': 'X23', 'F6': 'X18', 'F26': 'X4', 'F13': 'X7', 'F10': 'X17', 'F8': 'X3', 'F18': 'X22', 'F15': 'X5', 'F21': 'X9', 'F9': 'X12', 'F14': 'X19', 'F25': 'X11', 'F19': 'X16', 'F20': 'X14', 'F16': 'X21', 'F2': 'X20', 'F24': 'X13', 'F22': 'X6', 'F17': 'X26'}

{'F8': 'F3', 'F24': 'F1', 'F1': 'F5', 'F2': 'F4', 'F10': 'F7', 'F15': 'F11', 'F25': 'F23', 'F23': 'F12', 'F18': 'F6', 'F4': 'F26', 'F7': 'F13', 'F17': 'F10', 'F3': 'F8', 'F22': 'F18', 'F5': 'F15', 'F9': 'F21', 'F12': 'F9', 'F19': 'F14', 'F11': 'F25', 'F16': 'F19', 'F14': 'F20', 'F21': 'F16', 'F20': 'F2', 'F13': 'F24', 'F6': 'F22', 'F26': 'F17'}

{'C2': 'C2', 'C1': 'C1'}

Less

{'C2': 'Less', 'C1': 'More'}

RandomForestClassifier

C2

Student Job Placement

The classification algorithm's decision on the true label for the given case is solely dependent on the information presented to it. Per the algorithm, the accurate label for the case under consideration is most likely C2, and the 12.47% possibility of C1 reflects only a minor uncertainty in the classification algorithm's certainty. The marginal doubt mentioned above can be blamed on the negative contributions of F9, F1, F7, F4, and F3, supporting the assignment of C1 instead of C2. Conversely, the positive contributions of F8, F5, F11, F6, F2, and F12 are shifting the algorithm's decision higher in favour of label C2, hence the high certainty of its correctness. Overall, F9 and F1 are the most influential negative features, whereas F8 and F5 are the most positive features. Also, F10 is shown to have a negligible influence on the classification decision with respect to the case here.

{'C1': '12.47%', 'C2': '87.53%'}

{'F9': 'workex', 'F1': 'specialisation', 'F8': 'hsc_p', 'F5': 'gender', 'F11': 'mba_p', 'F7': 'hsc_s', 'F6': 'ssc_p', 'F2': 'etest_p', 'F4': 'ssc_b', 'F12': 'hsc_b', 'F3': 'degree_t', 'F10': 'degree_p'}

{'F11': 'F9', 'F12': 'F1', 'F2': 'F8', 'F6': 'F5', 'F5': 'F11', 'F9': 'F7', 'F1': 'F6', 'F4': 'F2', 'F7': 'F4', 'F8': 'F12', 'F10': 'F3', 'F3': 'F10'}

{'C2': 'C1', 'C1': 'C2'}

Placed

{'C1': 'Not Placed', 'C2': 'Placed'}

KNeighborsClassifier

C2

Printer Sales

The model indicates that the label for this case is likely C2, with an 83.33% chance that it is correct, implying that it is unlikely that C1 is the appropriate class. This predictive assertion is chiefly influenced by the values of the input variables F16, F18, and F17. While the F17 and F18 values positively control the model towards the prediction of C2, the F16 value biases the decision towards C1. However, the combined effect of F17 and F18 outweighs the contribution of F16. In addition, the variables F10, F19, and F2 also positively support the output predictions of the model. F1 has similar direction of contribution that of F16, further decreasing the odds of the C2 label. Unlike all the variables above, F6, F20, F11, F25, F15, and F22 are shown to have very little effect on model predictions with respect to the given case and we can say that their values receive very low consideration from the model.

{'C2': '83.33%', 'C1': '16.67%'}

{'F17': 'X24', 'F18': 'X1', 'F16': 'X4', 'F10': 'X10', 'F19': 'X2', 'F2': 'X8', 'F1': 'X17', 'F13': 'X7', 'F21': 'X21', 'F26': 'X18', 'F4': 'X6', 'F24': 'X11', 'F8': 'X22', 'F5': 'X25', 'F9': 'X5', 'F12': 'X19', 'F14': 'X15', 'F23': 'X23', 'F3': 'X16', 'F7': 'X3', 'F6': 'X14', 'F20': 'X20', 'F25': 'X13', 'F11': 'X12', 'F15': 'X9', 'F22': 'X26'}

{'F24': 'F17', 'F1': 'F18', 'F4': 'F16', 'F10': 'F10', 'F2': 'F19', 'F8': 'F2', 'F17': 'F1', 'F7': 'F13', 'F21': 'F21', 'F18': 'F26', 'F6': 'F4', 'F11': 'F24', 'F22': 'F8', 'F25': 'F5', 'F5': 'F9', 'F19': 'F12', 'F15': 'F14', 'F23': 'F23', 'F16': 'F3', 'F3': 'F7', 'F14': 'F6', 'F20': 'F20', 'F13': 'F25', 'F12': 'F11', 'F9': 'F15', 'F26': 'F22'}

{'C2': 'C2', 'C1': 'C1'}

Less

{'C2': 'Less', 'C1': 'More'}

LogisticRegression

C2

Hotel Satisfaction

The algorithm's forecast for the data instance under consideration is C2, and the decision's confidence level is about 91.36 percent. We can observe from the plot that the variables F12 and F13 are moving the prediction judgement towards the other label, C1. The F9, F7, F11, and F8, on the other hand, have values that have a favourable influence, pushing the data classification choice towards label C2. While F6 and F5 contradict the prediction, F14 and F2 have values that confirm the algorithm's prediction output verdict.

{'C2': '91.36%', 'C1': '8.64%'}

{'F12': 'Type of Travel', 'F13': 'Type Of Booking', 'F9': 'Hotel wifi service', 'F7': 'Common Room entertainment', 'F11': 'Stay comfort', 'F8': 'Other service', 'F6': 'Checkin\\/Checkout service', 'F14': 'Hotel location', 'F5': 'Food and drink', 'F2': 'Cleanliness', 'F1': 'Age', 'F3': 'Departure\\/Arrival convenience', 'F4': 'purpose_of_travel', 'F15': 'Ease of Online booking', 'F10': 'Gender'}

{'F3': 'F12', 'F4': 'F13', 'F6': 'F9', 'F12': 'F7', 'F11': 'F11', 'F14': 'F8', 'F13': 'F6', 'F9': 'F14', 'F10': 'F5', 'F15': 'F2', 'F5': 'F1', 'F7': 'F3', 'F2': 'F4', 'F8': 'F15', 'F1': 'F10'}

{'C2': 'C2', 'C1': 'C1'}

dissatisfied

{'C2': 'dissatisfied', 'C1': 'satisfied'}

SVC

C1

Vehicle Insurance Claims

The model classifies this case as C1 and it is noteworthy that there is, however, a 38.26% chance that the true label could be class C2. The uncertainty associated with the classification decision above is higher than expected, which could be attributed to the values of the different input features. The most influential feature is F1, which has a positive effect on the class C1 prediction by the model here. All other features are much less influential, with contributions from F8, F27, F15, and F9 shifting the prediction towards C2. Supporting the model in assigning the label choice, F10 is the next most influential feature. The impacts of the F2 and F32 are moderate, ranking seventh and eighth, respectively. Unfortunately, values of features such as F26, F29, F7, and F22 do not matter when determining the correct label in this instance.

{'C1': '61.74%', 'C2': '38.26%'}

{'F1': 'incident_severity', 'F8': 'insured_hobbies', 'F27': 'insured_occupation', 'F15': 'umbrella_limit', 'F9': 'policy_csl', 'F10': 'authorities_contacted', 'F2': 'insured_education_level', 'F32': 'collision_type', 'F19': 'months_as_customer', 'F31': 'vehicle_claim', 'F18': 'insured_relationship', 'F20': 'capital-gains', 'F23': 'auto_make', 'F13': 'injury_claim', 'F28': 'incident_city', 'F3': 'insured_sex', 'F25': 'number_of_vehicles_involved', 'F14': 'incident_hour_of_the_day', 'F24': 'age', 'F17': 'property_claim', 'F26': 'policy_annual_premium', 'F29': 'police_report_available', 'F7': 'property_damage', 'F22': 'incident_state', 'F6': 'policy_deductable', 'F5': 'capital-loss', 'F21': 'insured_zip', 'F16': 'incident_type', 'F12': 'bodily_injuries', 'F33': 'witnesses', 'F4': 'policy_state', 'F30': 'total_claim_amount', 'F11': 'auto_year'}

{'F27': 'F1', 'F23': 'F8', 'F22': 'F27', 'F5': 'F15', 'F19': 'F9', 'F28': 'F10', 'F21': 'F2', 'F26': 'F32', 'F1': 'F19', 'F16': 'F31', 'F24': 'F18', 'F7': 'F20', 'F33': 'F23', 'F14': 'F13', 'F30': 'F28', 'F20': 'F3', 'F10': 'F25', 'F9': 'F14', 'F2': 'F24', 'F15': 'F17', 'F4': 'F26', 'F32': 'F29', 'F31': 'F7', 'F29': 'F22', 'F3': 'F6', 'F8': 'F5', 'F6': 'F21', 'F25': 'F16', 'F11': 'F12', 'F12': 'F33', 'F18': 'F4', 'F13': 'F30', 'F17': 'F11'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Fraud', 'C2': 'Not Fraud'}

SVM_linear

C1

Wine Quality Prediction

The classification or prediction algorithm indicates that the most probable label for the given data is C1 since there is only a 25.47% chance that C2 could be the correct label. The major factors resulting in the above decision are F7, F9, and F1, while the set of features with moderate influence are F6, F3, F10, and F8. The least vital features are shown to be F5, F11, F2, and F4. In conclusion, it is very surprising to see the uncertainty surrounding the classification here given that only F6 and F10 have a negative impact, driving the algorithm to label the data as C2. To be specific, the contributions of F6 and F10 result in a decrease in the likelihood of C1 being the right label, as indicated by the prediction probabilities across the two possible classes but the influence of these negatives are moderated by the major positive features which are F7, F9, and F1.

{'C2': '25.47%', 'C1': '74.53%'}

{'F7': 'sulphates', 'F9': 'volatile acidity', 'F1': 'total sulfur dioxide', 'F6': 'residual sugar', 'F3': 'alcohol', 'F10': 'free sulfur dioxide', 'F8': 'chlorides', 'F5': 'fixed acidity', 'F11': 'citric acid', 'F2': 'pH', 'F4': 'density'}

{'F10': 'F7', 'F2': 'F9', 'F7': 'F1', 'F4': 'F6', 'F11': 'F3', 'F6': 'F10', 'F5': 'F8', 'F1': 'F5', 'F3': 'F11', 'F9': 'F2', 'F8': 'F4'}

{'C2': 'C2', 'C1': 'C1'}

high quality

{'C2': 'low_quality', 'C1': 'high quality'}

RandomForestClassifier

C1

Flight Price-Range Classification

The classification verdict is as follows: the most probable label for this case is C1, and the classifier is certain that neither C2 nor C3 is the correct label. The main drivers for the above classification are F12, F7, and F2, all of which have a strong positive influence, pushing the classifier to choose C1. Other positive features pushing the classification further higher towards C1 include F1, F8, F10, and F3. Not all the input features support the assigned label and the negative features F5, F6, and F9 indicate that the most probable class for this case could different from the assigned label. However, considering the confidence level in the above classification, it is valid to conclude that the classifier paid little attention to the negative features, hence selecting class C1.

{'C1': '100.00%', 'C2': '0.00%', 'C3': '0.00%'}

{'F12': 'Airline', 'F7': 'Duration_hours', 'F2': 'Total_Stops', 'F1': 'Journey_month', 'F6': 'Source', 'F8': 'Destination', 'F10': 'Arrival_hour', 'F5': 'Journey_day', 'F9': 'Dep_minute', 'F3': 'Arrival_minute', 'F11': 'Duration_mins', 'F4': 'Dep_hour'}

{'F9': 'F12', 'F7': 'F7', 'F12': 'F2', 'F2': 'F1', 'F10': 'F6', 'F11': 'F8', 'F5': 'F10', 'F1': 'F5', 'F4': 'F9', 'F6': 'F3', 'F8': 'F11', 'F3': 'F4'}

{'C1': 'C1', 'C2': 'C2', 'C3': 'C3'}

Low

{'C1': 'Low', 'C2': 'Moderate', 'C3': 'High'}

RandomForestClassifier

C1

Ethereum Fraud Detection

The best choice of label for the given case is C1 according to the classification algorithm, since there is little to no chance that C2 is the right class. Not all the features are shown to contribute either positively or negatively towards the label assigned here. The influential features can be ranked according to the associated degree of impact on the algorithm's output as follows: F4, F19, F9, F25, F31, F18, F33, F37, F3, F11, F15, F27, F12, F22, F1, F20, F35, F5, F16, F29. On the other hand, the irrelevant features include F17, F21, and F23 since they have close to zero impact. Among the top influential ones, F4, F19, F9, F25, and F31, the input feature F9 is regarded as the most negative, dragging the verdict in a different direction, while the others have positive contributions, improving the likelihood that the choice of C1 is appropriate in this case. The features with moderate influence are F33, F18, F37 where F33 is identified as a positive feature, while F18 and F37 considered negative features. Since a large number of top features have positive contributions that increase the probability that C1 is the right label, it is not surprising that the algorithm is very confident about the correctness of the assigned label.

{'C2': '0.00%', 'C1': '100.00%'}

{'F4': ' ERC20 total Ether sent contract', 'F19': ' ERC20 min val rec', 'F9': 'total transactions (including tnx to create contract', 'F25': ' ERC20 max val rec', 'F31': ' Total ERC20 tnxs', 'F18': ' ERC20 uniq rec addr', 'F33': 'min val sent', 'F37': 'Time Diff between first and last (Mins)', 'F3': 'Sent tnx', 'F11': 'Avg min between received tnx', 'F15': 'min value received', 'F27': ' ERC20 total ether sent', 'F12': 'avg val sent', 'F22': 'max val sent', 'F1': 'Avg min between sent tnx', 'F20': 'Received Tnx', 'F35': ' ERC20 uniq sent token name', 'F5': 'Unique Sent To Addresses', 'F16': ' ERC20 uniq rec token name', 'F29': ' ERC20 uniq rec contract addr', 'F17': 'total Ether sent', 'F21': 'Number of Created Contracts', 'F23': ' ERC20 avg val sent', 'F6': ' ERC20 max val sent', 'F32': ' ERC20 min val sent', 'F28': ' ERC20 avg val rec', 'F7': 'Unique Received From Addresses', 'F36': 'max value received ', 'F14': ' ERC20 uniq sent addr.1', 'F24': 'total ether sent contracts', 'F26': 'avg val received', 'F13': ' ERC20 uniq sent addr', 'F34': 'min value sent to contract', 'F10': 'max val sent to contract', 'F30': ' ERC20 total Ether received', 'F2': 'avg value sent to contract', 'F38': 'total ether balance', 'F8': 'total ether received'}

{'F26': 'F4', 'F31': 'F19', 'F18': 'F9', 'F32': 'F25', 'F23': 'F31', 'F28': 'F18', 'F12': 'F33', 'F3': 'F37', 'F4': 'F3', 'F2': 'F11', 'F9': 'F15', 'F25': 'F27', 'F14': 'F12', 'F13': 'F22', 'F1': 'F1', 'F5': 'F20', 'F37': 'F35', 'F8': 'F5', 'F38': 'F16', 'F30': 'F29', 'F19': 'F17', 'F6': 'F21', 'F36': 'F23', 'F35': 'F6', 'F34': 'F32', 'F33': 'F28', 'F7': 'F7', 'F10': 'F36', 'F29': 'F14', 'F21': 'F24', 'F11': 'F26', 'F27': 'F13', 'F15': 'F34', 'F16': 'F10', 'F24': 'F30', 'F17': 'F2', 'F22': 'F38', 'F20': 'F8'}

{'C1': 'C2', 'C2': 'C1'}

Fraud

{'C2': 'Not Fraud', 'C1': 'Fraud'}

LogisticRegression

C2

Student Job Placement

For the given case, the prediction decision is as follows: The probability of C1 being the correct label is only 18.57%, the probability of C2 is 81.43% making it the most probable label for the case here. The certainty of the prediction can be attributed to the influence of variables such as F11, F10, F5, F6, and F2. The least relevant variables considered to arrive at the classification verdict are F9, F3, F1, and F8. F7, F12, and F4 have moderate contributions to the classification here. The attribution analysis performed indicates that F6, F2, F12, F4, F3, and F8 are the negative variables, decreasing the likelihood of C2 in favour of labelling the given case as C1. The variables F11, F10, and F5 have the highest positive influence, which increases the odds of label C2 being the correct label.

{'C1': '18.57%', 'C2': '81.43%'}

{'F11': 'mba_p', 'F10': 'gender', 'F5': 'degree_t', 'F6': 'specialisation', 'F2': 'workex', 'F7': 'hsc_s', 'F12': 'hsc_p', 'F4': 'ssc_p', 'F9': 'ssc_b', 'F3': 'etest_p', 'F1': 'hsc_b', 'F8': 'degree_p'}

{'F5': 'F11', 'F6': 'F10', 'F10': 'F5', 'F12': 'F6', 'F11': 'F2', 'F9': 'F7', 'F2': 'F12', 'F1': 'F4', 'F7': 'F9', 'F4': 'F3', 'F8': 'F1', 'F3': 'F8'}

{'C2': 'C1', 'C1': 'C2'}

Placed

{'C1': 'Not Placed', 'C2': 'Placed'}

SGDClassifier

C1

Flight Price-Range Classification

The output decision of the classifier with respect to the given case is: C1 is the most probable label, followed by C3 and C2. To be specific, the predicted likelihood across the classes are as follows: 86.54% for C1, 13.46% for C3, and finally a 0.0% probability with respect to C2. The moderately high classification confidence could largely be due to the impact of certain input features supplied to the classifier. F6, F9, F2, F1, and F5 are the top-ranked variables whereas the least ranked are F3, F4, F12, F8, F10, F7, and F11. The marginal uncertainty in the classification verdict is due to the negative attributions of F9, F5, F12, F10, and F11 which prefer labelling the case differently. In conclusion, we can see that F6, F2, F1, F4, and F3 are among the positive variables pushing the classification in favour of C1.

{'C1': '86.54%', 'C3': '13.46%', 'C2': '0.00%'}

{'F6': 'Airline', 'F9': 'Total_Stops', 'F2': 'Source', 'F1': 'Journey_month', 'F5': 'Arrival_minute', 'F3': 'Journey_day', 'F4': 'Duration_hours', 'F12': 'Dep_hour', 'F8': 'Destination', 'F10': 'Arrival_hour', 'F7': 'Dep_minute', 'F11': 'Duration_mins'}

{'F9': 'F6', 'F12': 'F9', 'F10': 'F2', 'F2': 'F1', 'F6': 'F5', 'F1': 'F3', 'F7': 'F4', 'F3': 'F12', 'F11': 'F8', 'F5': 'F10', 'F4': 'F7', 'F8': 'F11'}

{'C1': 'C1', 'C2': 'C3', 'C3': 'C2'}

Low

{'C1': 'Low', 'C3': 'Moderate', 'C2': 'High'}

LogisticRegression

C1

Airline Passenger Satisfaction

C1 is the predicted label assigned to this case or instance. This is based on the fact that there is only a 0.68% chance that C2 is the correct label. The most relevant variables that increase the prediction's probability are F11, F5, F9, and F19. Conversely, F2 is the only important feature driving the classification decision in the direction of C2. Other negative features include F12, F18, F8, and F17. Other positive features increasing the chances of the C1 prediction are F3, F16, and F1. Unlike F11, F5, F9, and F19, these positive variables have moderate contributions to the model's decision. The least ranked among all the relevant features are F7, F6, F22, and F20, with lower attributions to the C1 prediction, however, F10 and F4 are shown to have no impact when determining the correct label for the case under consideration.

{'C1': '99.32%', 'C2': '0.68%'}

{'F11': 'Type of Travel', 'F2': 'Customer Type', 'F5': 'Inflight entertainment', 'F19': 'Inflight wifi service', 'F9': 'Departure\\/Arrival time convenient', 'F12': 'Gate location', 'F18': 'Arrival Delay in Minutes', 'F8': 'Seat comfort', 'F17': 'Online boarding', 'F3': 'Ease of Online booking', 'F16': 'Class', 'F1': 'Age', 'F15': 'On-board service', 'F14': 'Cleanliness', 'F13': 'Checkin service', 'F21': 'Inflight service', 'F7': 'Food and drink', 'F6': 'Departure Delay in Minutes', 'F22': 'Baggage handling', 'F20': 'Gender', 'F10': 'Flight Distance', 'F4': 'Leg room service'}

{'F4': 'F11', 'F2': 'F2', 'F14': 'F5', 'F7': 'F19', 'F8': 'F9', 'F10': 'F12', 'F22': 'F18', 'F13': 'F8', 'F12': 'F17', 'F9': 'F3', 'F5': 'F16', 'F3': 'F1', 'F15': 'F15', 'F20': 'F14', 'F18': 'F13', 'F19': 'F21', 'F11': 'F7', 'F21': 'F6', 'F17': 'F22', 'F1': 'F20', 'F6': 'F10', 'F16': 'F4'}

{'C2': 'C1', 'C1': 'C2'}

neutral or dissatisfied

{'C1': 'neutral or dissatisfied', 'C2': 'satisfied'}

RandomForestClassifier

C1

Personal Loan Modelling

The following classification decisions are largely based on the factors or attributes of this particular case. The class label, in this case, is projected to be C1 out of the potential classes, which is 97.50% likely. The next possible label is C2, which has an approximate probability of 2.50%. The confidence level with respect to this classification is very high, and the features with the most contributions are F4, F6, and F2. However, F7, F9, and F3 are shown to be the least relevant features. The attribution analysis shows that the only positive features whose contributions favour labelling the case as C1 are F6, F2, F8, and F5. However, the negative attributions of F4, F1, F9, F7, and F3 also indicate that perhaps C2 could be the true label. Judging based on the confidence level coupled with the attributions, it can be concluded that the values of the positive features F6, F2, F8, and F5 are good enough to steer the classification in the direction of C1, but the strong negative attribution of F4 casts about 2.50% of doubt on the decision.

{'C1': '97.50%', 'C2': '2.50%'}

{'F4': 'Income', 'F6': 'CD Account', 'F2': 'Education', 'F1': 'Securities Account', 'F8': 'CCAvg', 'F5': 'Family', 'F9': 'Extra_service', 'F7': 'Age', 'F3': 'Mortgage'}

{'F2': 'F4', 'F8': 'F6', 'F5': 'F2', 'F7': 'F1', 'F4': 'F8', 'F3': 'F5', 'F9': 'F9', 'F1': 'F7', 'F6': 'F3'}

{'C1': 'C1', 'C2': 'C2'}

Reject

{'C1': 'Reject', 'C2': 'Accept'}

DNN

C2

Ethereum Fraud Detection

The prediction probabilities for classes C1 and C2, respectively, are 15.35% and 84.65%. Based on the aforementioned, C2 is the most likely class label for the presented data instance, and according to the attribution analysis, the various input variables had varying degrees of impact on the model's classification judgement. F20, F30, F37, F4, F23, F1, and F28 are the most influential factors, whereas F10, F12, F35, F7, F11, and F19 have the least impact. The subsequent analysis will concentrate on the most relevant factors influencing the label selection in this case. Looking at the attributions of the input features, only F20 and F30 exhibit negative contributions among the top influential features, F20, F30, F37, F4, and F28, lowering the chance that C2 is the right label, and they strongly favour labelling the instance as C1 instead. Positive variables such as F37, F4, and F28 influence the classification choice in favour of C2. The remaining variables, including F23, F1, and F36, have a moderate to low impact. In essence, the marginal uncertainty in this decision is mostly owing to the negative impacts of F20, F30, F26, and F13, while the positive contributions of F37, F4, F36, F23, F1, and F28 push the decision much closer to C2.

{'C1': '15.35%', 'C2': '84.65%'}

{'F20': ' ERC20 uniq rec contract addr', 'F30': ' ERC20 uniq rec token name', 'F37': 'min value received', 'F4': 'Time Diff between first and last (Mins)', 'F28': 'avg val sent', 'F23': ' ERC20 uniq sent token name', 'F1': 'Sent tnx', 'F36': 'Avg min between received tnx', 'F26': 'Unique Received From Addresses', 'F13': ' ERC20 uniq rec addr', 'F16': 'total transactions (including tnx to create contract', 'F5': 'Avg min between sent tnx', 'F33': ' ERC20 uniq sent addr.1', 'F38': 'avg val received', 'F6': 'Unique Sent To Addresses', 'F18': 'max value received ', 'F27': 'max val sent', 'F3': 'min val sent', 'F9': 'Number of Created Contracts', 'F24': 'total ether received', 'F29': ' ERC20 uniq sent addr', 'F25': ' ERC20 total Ether received', 'F2': 'Received Tnx', 'F34': ' ERC20 avg val sent', 'F32': 'total Ether sent', 'F21': ' ERC20 min val sent', 'F15': 'max val sent to contract', 'F17': 'total ether balance', 'F22': ' ERC20 max val sent', 'F8': ' Total ERC20 tnxs', 'F31': ' ERC20 total ether sent', 'F14': ' ERC20 avg val rec', 'F35': 'avg value sent to contract', 'F12': ' ERC20 min val rec', 'F10': ' ERC20 max val rec', 'F19': ' ERC20 total Ether sent contract', 'F11': 'min value sent to contract', 'F7': 'total ether sent contracts'}

{'F30': 'F20', 'F38': 'F30', 'F9': 'F37', 'F3': 'F4', 'F14': 'F28', 'F37': 'F23', 'F4': 'F1', 'F2': 'F36', 'F7': 'F26', 'F28': 'F13', 'F18': 'F16', 'F1': 'F5', 'F29': 'F33', 'F11': 'F38', 'F8': 'F6', 'F10': 'F18', 'F13': 'F27', 'F12': 'F3', 'F6': 'F9', 'F20': 'F24', 'F27': 'F29', 'F24': 'F25', 'F5': 'F2', 'F36': 'F34', 'F19': 'F32', 'F34': 'F21', 'F16': 'F15', 'F22': 'F17', 'F35': 'F22', 'F23': 'F8', 'F25': 'F31', 'F33': 'F14', 'F17': 'F35', 'F31': 'F12', 'F32': 'F10', 'F26': 'F19', 'F15': 'F11', 'F21': 'F7'}

{'C2': 'C1', 'C1': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

RandomForestClassifier

C2

Company Bankruptcy Prediction

The model assigns the class C2 with near perfect certainty or confidence level since the predicted likelihood of C1 is only 1.0%. F10, F17, F1, F16, and F36 have the greatest cumulative beneficial influence on the model's choice to create C2. F22 also had a significant influence, but it shifted the choice away from C2. Furthermore, F85 and F51 had a modest influence on C2 decision making, which was still bigger than features F18 and F86, which had a moderate impact and contributed to C1 class prediction. Furthermore, F51, F54, and F29 have minimal positive impact on the final result, further increasing the chances of C2 being the appropriate label for the given case. However, a number of input features, notably F64, F37, F4, and F15, appear to be less essential to predictions here. All in all, the very high confidence level could easily be explained away by considering the fact that the joint influence of the positive variables such as F10, F17, F1, F16, and F36 far outshines the joint contribution of the negative variables such as F22, F18, F86, and F53.

{'C2': '99.00%', 'C1': '1.00%'}

{'F10': " Net Income to Stockholder's Equity", 'F1': ' Continuous interest rate (after tax)', 'F16': ' ROA(C) before interest and depreciation before interest', 'F22': ' Borrowing dependency', 'F17': ' Cash Flow Per Share', 'F36': ' Net worth\\/Assets', 'F85': ' Total income\\/Total expense', 'F51': ' Persistent EPS in the Last Four Seasons', 'F54': ' Retained Earnings to Total Assets', 'F29': ' Net Value Per Share (B)', 'F35': ' Cash Flow to Equity', 'F80': ' Net Value Per Share (A)', 'F65': ' Degree of Financial Leverage (DFL)', 'F18': ' Per Share Net profit before tax (Yuan ¥)', 'F86': ' Revenue Per Share (Yuan ¥)', 'F53': ' Inventory Turnover Rate (times)', 'F14': ' Net profit before tax\\/Paid-in capital', 'F77': ' Equity to Long-term Liability', 'F6': ' Operating profit\\/Paid-in capital', 'F69': ' Cash Turnover Rate', 'F64': ' Operating Funds to Liability', 'F37': ' Contingent liabilities\\/Net worth', 'F4': ' Working Capital to Total Assets', 'F15': ' Liability to Equity', 'F11': ' Current Liability to Liability', 'F34': ' Operating Gross Margin', 'F39': ' Operating Profit Per Share (Yuan ¥)', 'F42': ' Long-term Liability to Current Assets', 'F23': ' Current Asset Turnover Rate', 'F43': ' Interest Coverage Ratio (Interest expense to EBIT)', 'F78': ' Equity to Liability', 'F81': ' Operating Profit Rate', 'F67': ' Current Liability to Equity', 'F72': ' No-credit Interval', 'F58': ' Net Worth Turnover Rate (times)', 'F55': ' Working Capital\\/Equity', 'F41': ' Quick Assets\\/Current Liability', 'F90': ' Inventory and accounts receivable\\/Net value', 'F5': ' Current Liability to Current Assets', 'F68': ' Working capitcal Turnover Rate', 'F45': ' Fixed Assets to Assets', 'F33': ' Continuous Net Profit Growth Rate', 'F27': ' Cash Reinvestment %', 'F20': ' CFO to Assets', 'F83': ' Total Asset Turnover', 'F31': ' After-tax net Interest Rate', 'F2': ' After-tax Net Profit Growth Rate', 'F50': ' Tax rate (A)', 'F59': ' Current Ratio', 'F47': ' Realized Sales Gross Margin', 'F92': ' Net Value Per Share (C)', 'F57': ' Regular Net Profit Growth Rate', 'F84': ' Interest-bearing debt interest rate', 'F44': ' Debt ratio %', 'F32': ' Long-term fund suitability ratio (A)', 'F24': ' Net Value Growth Rate', 'F25': ' Total Asset Growth Rate', 'F9': ' Fixed Assets Turnover Frequency', 'F12': ' Inventory\\/Current Liability', 'F62': ' Allocation rate per person', 'F30': ' Operating Expense Rate', 'F48': ' Operating profit per person', 'F28': ' Net Income to Total Assets', 'F74': ' Interest Expense Ratio', 'F61': ' Cash\\/Total Assets', 'F40': ' ROA(B) before interest and depreciation after tax', 'F89': ' Inventory\\/Working Capital', 'F52': ' Total assets to GNP price', 'F3': ' Total debt\\/Total net worth', 'F21': ' Quick Ratio', 'F82': ' Revenue per person', 'F93': ' Non-industry income and expenditure\\/revenue', 'F63': ' Cash Flow to Sales', 'F8': ' ROA(A) before interest and % after tax', 'F26': ' Current Liabilities\\/Liability', 'F73': ' Operating Profit Growth Rate', 'F49': ' Cash Flow to Liability', 'F7': ' Cash Flow to Total Assets', 'F79': ' Pre-tax net Interest Rate', 'F70': ' Accounts Receivable Turnover', 'F56': ' Current Liability to Assets', 'F19': ' Quick Assets\\/Total Assets', 'F60': ' Total expense\\/Assets', 'F75': ' Average Collection Days', 'F76': ' Research and development expense rate', 'F38': ' Current Assets\\/Total Assets', 'F88': ' Current Liabilities\\/Equity', 'F87': ' Realized Sales Gross Profit Growth Rate', 'F46': ' Cash flow rate', 'F66': ' Total Asset Return Growth Rate Ratio', 'F13': ' Quick Asset Turnover Rate', 'F91': ' Cash\\/Current Liability', 'F71': ' Gross Profit to Sales'}

{'F59': 'F10', 'F12': 'F1', 'F29': 'F16', 'F3': 'F22', 'F65': 'F17', 'F84': 'F36', 'F57': 'F85', 'F8': 'F51', 'F10': 'F54', 'F27': 'F29', 'F53': 'F35', 'F42': 'F80', 'F35': 'F65', 'F78': 'F18', 'F31': 'F86', 'F18': 'F53', 'F72': 'F14', 'F23': 'F77', 'F89': 'F6', 'F34': 'F69', 'F87': 'F64', 'F64': 'F37', 'F67': 'F4', 'F66': 'F15', 'F90': 'F11', 'F62': 'F34', 'F63': 'F39', 'F69': 'F42', 'F61': 'F23', 'F60': 'F43', 'F91': 'F78', 'F58': 'F81', 'F92': 'F67', 'F56': 'F72', 'F55': 'F58', 'F68': 'F55', 'F71': 'F41', 'F70': 'F90', 'F86': 'F5', 'F73': 'F68', 'F74': 'F45', 'F54': 'F33', 'F75': 'F27', 'F76': 'F20', 'F77': 'F83', 'F79': 'F31', 'F80': 'F2', 'F81': 'F50', 'F82': 'F59', 'F83': 'F47', 'F88': 'F92', 'F85': 'F57', 'F1': 'F84', 'F47': 'F44', 'F52': 'F32', 'F15': 'F24', 'F24': 'F25', 'F22': 'F9', 'F21': 'F12', 'F20': 'F62', 'F19': 'F30', 'F17': 'F48', 'F16': 'F28', 'F14': 'F74', 'F26': 'F61', 'F13': 'F40', 'F11': 'F89', 'F9': 'F52', 'F7': 'F3', 'F6': 'F21', 'F5': 'F82', 'F4': 'F93', 'F25': 'F63', 'F28': 'F8', 'F51': 'F26', 'F43': 'F73', 'F50': 'F49', 'F49': 'F7', 'F48': 'F79', 'F2': 'F70', 'F46': 'F56', 'F45': 'F19', 'F44': 'F60', 'F41': 'F75', 'F30': 'F76', 'F40': 'F38', 'F39': 'F88', 'F38': 'F87', 'F37': 'F46', 'F36': 'F66', 'F33': 'F13', 'F32': 'F91', 'F93': 'F71'}

{'C1': 'C2', 'C2': 'C1'}

No

{'C2': 'No', 'C1': 'Yes'}

RandomForestClassifier

C2

House Price Classification

Between the two classes, the model labelled this case as C2 with a likelihood of about 97.0% since there is only a marginal chance that it belongs to label C1. The most relevant features influencing this decision are F10, F13, F11, and F2. In this case, F10, F13, and F2 have a considerable positive influence on the prediction of C2. In contrast, the values of F11 and F3 throw a bit of doubt on the C2 prediction. However, compared to F10, F13, and F2, this shift is very small. Finally, there are some attributes with limited impact on the prediction of C2 and these are F4, F5, F8, F12, F1, and F9 since their values are less important to the model in terms of determining the label for this case.

{'C1': '3.00%', 'C2': '97.00%'}

{'F10': 'LSTAT', 'F13': 'RM', 'F2': 'AGE', 'F11': 'TAX', 'F7': 'PTRATIO', 'F3': 'DIS', 'F6': 'CRIM', 'F4': 'RAD', 'F5': 'B', 'F8': 'NOX', 'F12': 'ZN', 'F1': 'INDUS', 'F9': 'CHAS'}

{'F13': 'F10', 'F6': 'F13', 'F7': 'F2', 'F10': 'F11', 'F11': 'F7', 'F8': 'F3', 'F1': 'F6', 'F9': 'F4', 'F12': 'F5', 'F5': 'F8', 'F2': 'F12', 'F3': 'F1', 'F4': 'F9'}

{'C2': 'C1', 'C1': 'C2'}

High

{'C1': 'Low', 'C2': 'High'}

LogisticRegression

C1

Concrete Strength Classification

Probably C1 is the right label for this case since the probability of the alternative label, C2 and C3, are only 1.03% and 0.0%. The order of importance of the features for the above classification verdict is F6, F2, F8, F4, F5, F3, F1, and F7. Analysis conducted shows that only the features F2, F5, and F3 have negative contributions, hence reducing the probability of assigning label C1 to the given case. Positive features that increase the likelihood that C1 is the valid label are F6, F8, F4, F1, and F7. The co-attribution of the positive variables is stronger than that of the negative ones, so it is not surprising that we see the level of confidence associated with the prediction of class C1.

{'C2': '1.03%', 'C1': '98.97%', 'C3': '0.0%'}

{'F6': 'cement', 'F2': 'age_days', 'F8': 'water', 'F4': 'superplasticizer', 'F5': 'fineaggregate', 'F3': 'flyash', 'F1': 'slag', 'F7': 'coarseaggregate'}

{'F1': 'F6', 'F8': 'F2', 'F4': 'F8', 'F5': 'F4', 'F7': 'F5', 'F3': 'F3', 'F2': 'F1', 'F6': 'F7'}

{'C3': 'C2', 'C1': 'C1', 'C2': 'C3'}

Strong

{'C2': 'Weak', 'C1': 'Strong', 'C3': 'Other'}

DNN

C2

Credit Card Fraud Classification

The model labels the given data as C2 since it has a higher predicted probability equal to 51.42% compared to that of C1 which is equal to 48.58%. The input variables with higher contributions to the above classification decision are F22, F14, F2, F19, and F9, while those with little influence are F8, F13, F30, F1, and F23. Positively supporting the choice of the label, in this case, are mainly F22, F14, F9, and F19. However, the main negative variables are F2, F4, and F10. Judging based on the degree of influence as well as the direction of influence of the variables, it is not surprising that the model is only 51.42% confident in the assigned label which is marginally above average.

{'C1': '48.58%', 'C2': '51.42%'}

{'F14': 'Z18', 'F22': 'Z14', 'F2': 'Time', 'F19': 'Z1', 'F9': 'Z19', 'F20': 'Z10', 'F5': 'Z4', 'F3': 'Z3', 'F18': 'Z12', 'F17': 'Z16', 'F4': 'Z7', 'F10': 'Z11', 'F29': 'Z9', 'F6': 'Z6', 'F28': 'Z23', 'F16': 'Z5', 'F15': 'Z17', 'F11': 'Z21', 'F12': 'Z24', 'F24': 'Z8', 'F26': 'Amount', 'F25': 'Z20', 'F7': 'Z27', 'F27': 'Z25', 'F21': 'Z13', 'F8': 'Z2', 'F13': 'Z22', 'F30': 'Z28', 'F1': 'Z26', 'F23': 'Z15'}

{'F19': 'F14', 'F15': 'F22', 'F1': 'F2', 'F2': 'F19', 'F20': 'F9', 'F11': 'F20', 'F5': 'F5', 'F4': 'F3', 'F13': 'F18', 'F17': 'F17', 'F8': 'F4', 'F12': 'F10', 'F10': 'F29', 'F7': 'F6', 'F24': 'F28', 'F6': 'F16', 'F18': 'F15', 'F22': 'F11', 'F25': 'F12', 'F9': 'F24', 'F30': 'F26', 'F21': 'F25', 'F28': 'F7', 'F26': 'F27', 'F14': 'F21', 'F3': 'F8', 'F23': 'F13', 'F29': 'F30', 'F27': 'F1', 'F16': 'F23'}

{'C1': 'C1', 'C2': 'C2'}

Fraud

{'C1': 'Not Fraud', 'C2': 'Fraud'}

SGDClassifier

C2

Company Bankruptcy Prediction

The following is the classification for the provided data: C2 is the most likely class label and C1 cannot possibly be the correct label given the likelihood is 0.0%. F73, F34, and F8 are the key variables that contributed to the classification choice. However, the classifier does not consider all features while making this conclusion, and these irrelevant features include F43, F39, F10, and F91. Revealed to have positive contributions to the prediction made here among the top features are F8, F5, F27, and F32, but all of the others, F73, F34, F36, F20, F51, F12, and F46, argue against labelling the present scenario as C1 and despite the fact that the bulk of relevant features are pointing in the opposite direction, the classifier is extremely certain that the proper label for the current scenario is C2, not C1.

{'C1': '0.00%', 'C2': '100.00%'}

{'F73': ' Liability to Equity', 'F34': ' Net worth\\/Assets', 'F8': ' Debt ratio %', 'F36': " Net Income to Stockholder's Equity", 'F46': ' Equity to Liability', 'F32': ' Realized Sales Gross Margin', 'F20': ' Net Value Per Share (A)', 'F51': ' Current Liability to Assets', 'F5': ' Current Liability to Equity', 'F27': ' Net Income to Total Assets', 'F12': ' Operating Profit Per Share (Yuan ¥)', 'F19': ' ROA(B) before interest and depreciation after tax', 'F83': ' Working Capital to Total Assets', 'F1': ' Persistent EPS in the Last Four Seasons', 'F68': ' Current Liabilities\\/Equity', 'F89': ' Total expense\\/Assets', 'F15': ' Net Value Per Share (C)', 'F6': ' Gross Profit to Sales', 'F61': ' Pre-tax net Interest Rate', 'F81': ' Cash\\/Current Liability', 'F43': ' Total assets to GNP price', 'F39': ' Working capitcal Turnover Rate', 'F10': ' Net profit before tax\\/Paid-in capital', 'F91': ' Quick Assets\\/Current Liability', 'F33': ' Inventory and accounts receivable\\/Net value', 'F55': ' Long-term Liability to Current Assets', 'F41': ' Working Capital\\/Equity', 'F70': ' Operating Expense Rate', 'F30': ' Cash Reinvestment %', 'F93': ' Retained Earnings to Total Assets', 'F84': ' Cash Flow Per Share', 'F22': ' Contingent liabilities\\/Net worth', 'F66': ' Inventory\\/Working Capital', 'F74': ' Operating Gross Margin', 'F82': ' Current Asset Turnover Rate', 'F2': ' Interest Coverage Ratio (Interest expense to EBIT)', 'F88': ' Fixed Assets to Assets', 'F90': ' CFO to Assets', 'F54': ' Operating Profit Rate', 'F69': ' Total Asset Turnover', 'F80': ' Borrowing dependency', 'F62': ' Non-industry income and expenditure\\/revenue', 'F37': ' Current Liability to Liability', 'F77': ' Operating profit\\/Paid-in capital', 'F60': ' Revenue per person', 'F52': ' Operating Funds to Liability', 'F23': ' Current Liability to Current Assets', 'F18': ' Regular Net Profit Growth Rate', 'F64': ' Quick Ratio', 'F21': ' Total debt\\/Total net worth', 'F47': ' Current Ratio', 'F40': ' Tax rate (A)', 'F56': ' After-tax Net Profit Growth Rate', 'F9': ' After-tax net Interest Rate', 'F72': ' Per Share Net profit before tax (Yuan ¥)', 'F26': ' Continuous interest rate (after tax)', 'F31': ' No-credit Interval', 'F79': ' Total income\\/Total expense', 'F86': ' Allocation rate per person', 'F35': ' Total Asset Return Growth Rate Ratio', 'F50': ' Degree of Financial Leverage (DFL)', 'F25': ' Cash Turnover Rate', 'F65': ' Quick Asset Turnover Rate', 'F13': ' Revenue Per Share (Yuan ¥)', 'F44': ' Research and development expense rate', 'F75': ' ROA(C) before interest and depreciation before interest', 'F14': ' ROA(A) before interest and % after tax', 'F11': ' Net Value Per Share (B)', 'F7': ' Cash\\/Total Assets', 'F4': ' Cash Flow to Sales', 'F58': ' Total Asset Growth Rate', 'F16': ' Equity to Long-term Liability', 'F67': ' Fixed Assets Turnover Frequency', 'F17': ' Inventory\\/Current Liability', 'F87': ' Cash flow rate', 'F28': ' Realized Sales Gross Profit Growth Rate', 'F71': ' Inventory Turnover Rate (times)', 'F49': ' Cash Flow to Total Assets', 'F3': ' Net Worth Turnover Rate (times)', 'F48': ' Continuous Net Profit Growth Rate', 'F78': ' Cash Flow to Equity', 'F53': ' Long-term fund suitability ratio (A)', 'F42': ' Current Liabilities\\/Liability', 'F59': ' Cash Flow to Liability', 'F57': ' Accounts Receivable Turnover', 'F29': ' Current Assets\\/Total Assets', 'F85': ' Interest Expense Ratio', 'F45': ' Quick Assets\\/Total Assets', 'F76': ' Net Value Growth Rate', 'F24': ' Operating Profit Growth Rate', 'F63': ' Operating profit per person', 'F38': ' Average Collection Days', 'F92': ' Interest-bearing debt interest rate'}

{'F66': 'F73', 'F84': 'F34', 'F47': 'F8', 'F59': 'F36', 'F91': 'F46', 'F83': 'F32', 'F42': 'F20', 'F46': 'F51', 'F92': 'F5', 'F16': 'F27', 'F63': 'F12', 'F13': 'F19', 'F67': 'F83', 'F8': 'F1', 'F39': 'F68', 'F44': 'F89', 'F88': 'F15', 'F93': 'F6', 'F48': 'F61', 'F32': 'F81', 'F9': 'F43', 'F73': 'F39', 'F72': 'F10', 'F71': 'F91', 'F70': 'F33', 'F69': 'F55', 'F68': 'F41', 'F19': 'F70', 'F75': 'F30', 'F10': 'F93', 'F65': 'F84', 'F64': 'F22', 'F11': 'F66', 'F62': 'F74', 'F61': 'F82', 'F60': 'F2', 'F74': 'F88', 'F76': 'F90', 'F58': 'F54', 'F77': 'F69', 'F3': 'F80', 'F4': 'F62', 'F90': 'F37', 'F89': 'F77', 'F5': 'F60', 'F87': 'F52', 'F86': 'F23', 'F85': 'F18', 'F6': 'F64', 'F7': 'F21', 'F82': 'F47', 'F81': 'F40', 'F80': 'F56', 'F79': 'F9', 'F78': 'F72', 'F12': 'F26', 'F56': 'F31', 'F57': 'F79', 'F20': 'F86', 'F36': 'F35', 'F35': 'F50', 'F34': 'F25', 'F33': 'F65', 'F31': 'F13', 'F30': 'F44', 'F29': 'F75', 'F28': 'F14', 'F27': 'F11', 'F26': 'F7', 'F25': 'F4', 'F24': 'F58', 'F23': 'F16', 'F22': 'F67', 'F21': 'F17', 'F37': 'F87', 'F38': 'F28', 'F18': 'F71', 'F49': 'F49', 'F55': 'F3', 'F54': 'F48', 'F53': 'F78', 'F52': 'F53', 'F51': 'F42', 'F50': 'F59', 'F2': 'F57', 'F40': 'F29', 'F14': 'F85', 'F45': 'F45', 'F15': 'F76', 'F43': 'F24', 'F17': 'F63', 'F41': 'F38', 'F1': 'F92'}

{'C2': 'C1', 'C1': 'C2'}

Yes

{'C1': 'No', 'C2': 'Yes'}