Datasets:

james-burton

/

aug-text-exps-v3

like 0

KNeighborsClassifier

C2

Cab Surge Pricing System

The correct label, according to the classifier, is neither C3 nor C1, but C2, with a prediction likelihood of about 75.0%. By analysing the attributions of the input features, they can be ranked according to the level of impact, from the most important feature to the least relevant, as follows: F4, F8, F5, F9, F6, F1, F10, F7, F2, F3, F11, and F12. Among the twelve features considered by the classifier for the prediction verdict, seven have a positive influence on the classifier. F5, F9, F10, F1, and F7 are the five negative features that swing the assessment decision towards other classes. The value of F4 has a strong positive contribution to increasing classifier's response, favouring the assigning of C2. The last four features, F2, F3, F11, and F12, have a weak positive effect on the classifier's prediction for this case.

{'C3': '25.00%', 'C2': '75.00%', 'C1': '0.00%'}

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

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

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

C2

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

KNeighborsClassifier

C1

Basketball Players Career Length Prediction

It is important to note that the classifier's labelling decision is based solely on the information supplied. The classification verdict is as follows: C1 is the most probable label with respect to the case under consideration, since the prediction likelihood of the other label, C2, is only 12.50%. The most important variables contributing to the abovementioned classification are F5, F14, and F10, whereas remaining variables such as F17, F15, F1, F11, and F13 have a modest effect on the classifier's labelling decision for the given case. All the top features positively support the selection of C1 as the correct label and the negative variables increasing the chances of C2 are F19, F9, and F12. Given that these are the variables reducing the classifier's response towards generating label C1, it is not surprising that the classifier is very confident that C1 is likely the true label. In addition, the joint negative attribution of F19, F9, and F12 is very small when compared with the positive attributions of F5, F10, F17, and F14.

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

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

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

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

More than 5

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

DecisionTreeClassifier

C1

Concrete Strength Classification

According to the classification algorithm or model, C1 is the most likely class, with a very high confidence level, and C2 has a very low likelihood of being the right label. All of the inputs are proven to contribute to the categorization described above and the following is a ordering of the features from least essential to most significant based on their degree of influence: F8, F3, F4, F7, F5, F1, F6, and F2. It is clear from the attributions of the input attributes that the algorithm is quite certain that C2 is not the proper label for the given case since each attribute contributes positively, resulting in a significant push towards C1.

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

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

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

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

Weak

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

RandomForestClassifier

C2

Mobile Price-Range Classification

The model indicates that C4 and C3 have zero prediction probabilities, while that of C1 is 3.85%, meaning the most probable label for the given case is C2 and the confidence level is approximately equal to 96.15% certainty. The major features driving the above classification are F7, F9, and F8, while the least relevant features are F5, F16, F3, F14, and F2. The intermediate features have varying degrees of influence, from moderate to low, and these include F10, F6, and F11. Among the top influential features, only F10 has a negative contribution, driving the prediction slightly towards one of the other possible classes. Furthermore, the top two positive features, F9 and F7, have a stronger influence than all the negative features combined. It is, therefore, not surprising that the model is confident about the classification verdict here.

{'C4': '0.00%', 'C3': '0.00%', 'C1': '3.85%', 'C2': '96.15%'}

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

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

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

r4

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

SVM_poly

C1

Mobile Price-Range Classification

The predicted output label from the model is C1 with almost 100% certainty, indicating it is very certain it is correct and this is mainly because the likelihoods across the other labels C2, C4, and C3 are 0.47%, 0.05%, and 0.04%, respectively. Among the top features F3, F9, and F14, the features F14 and F9 positively influence the classification decision above in the direction of C1, whereas F3 influences in the opposite direction in favour of an alternative label. With a similar direction of influence as F3, the features F1, F13, F4, and F2 negatively impact the prediction of C1, whereas F15 positively impacts it. Features F7, F13, F17, and F10 also have a smaller influence on the prediction output for the given case and finally, the features F11, F20, and F19, have very little contributions to the classification made by the model for the case under consideration.

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

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

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

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

r1

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

SVC

C1

Paris House Classification

The prediction probabilities associated with the classes C1 and C2 are 99.56% and 0.44%, respectively. Therefore, we can conclude that the most probable label for the given data is C1. The classification model's decision here is largely based on the impacts of the F5, F16, and F9, whereas the F14, F1, and F13 have very little to say about the decision here. In terms of the direction of influence of the features, F5, F4, F10, F8, and F6 are the top positive features contributing to the prediction outcome of C1. Conversely, the marginal doubt in the classification decision (represented by the probability of C2) is largely due to the negative contributions of F16, F11, F12, and F2. To sum up, the very high certainty in the classification output decision could be explained by considering the fact that the joint influence of the negative features is smaller than that of the positive features.

{'C1': '99.56%', 'C2': '0.44%'}

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

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

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

Basic

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

GradientBoostingClassifier

C1

Basketball Players Career Length Prediction

The model identifies the case as C1 since, the true label has just 33.63 percent chance of being C2 when the prediction probability is calculated. The in-depth analysis found that the bulk of the attributes had negative impacts, driving the prediction away from C1 and toward C2. F13, F2, F7, F9, and F17 are among the features that contribute negatively. Furthermore, these features' values are ranked higher than any of the positive features, which are F15, F5, F4, and F12. Finally, it can be concluded that the values of F3, F1, and F8 are less important in predicting the outcome of the case under review, hence they are ranked the least.

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

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

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

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

Less than 5

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

KNeighborsClassifier

C1

Cab Surge Pricing System

0.0% is the predicted probability that C3 is the true label for the test example under consideration according to the classifier. Judging based on the predicted probabilities associated with the other remaining labels, the classifier is 75.0% confident that C1 is the correct label. From the analysis, the features ranked according to the degree of impact from the most significant feature to the least relevant ones: F12, F6, F3, F4, F5, F11, F8, F2, F1, F7, F9, and F10. Examining the contributions or attributions of the features further revealed that the ratio of positive features to negative features is seven to five. The negative features swinging the prediction decision towards the other classes are F3, F4, F8, F11, and F2 since their contribution decrease the probability that C1 is the true label for the given case. The value of F12 has the strongest positive contribution increasing the classifier's response in support of assigning C1 but the last four features, F1, F7, F9, and F10, have a weak positive influence on the labelling decision or conclusion with respect to the given case.

{'C2': '25.00%', 'C1': '75.00%', 'C3': '0.00%'}

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

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

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

C2

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

SGDClassifier

C2

House Price Classification

The classifier is very certain that C1 is not the true label since the predicted probability of C2 is given as 100.0%. Analysing the attributions of the features indicates that the most relevant features are F4, F8, F12, and F10 while F11, F6, and F13 are the least relevant features. The values of F5, F2, F1, F3, F9, and F7 have a moderate influence on the classification decision made here. Considering that the classifier is 100.0% certain that C2 is the true label, we can conclude that the collective negative attribution of F12, F2, and F7 is clearly outweighed by the positive attributions of features such as F4, F8, F5, and F10.

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

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

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

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

Low

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

GradientBoostingClassifier

C1

Health Care Services Satisfaction Prediction

Based on the information provided to the classifier, the true label for the given case is likely C1, with a confidence level of 76.26%. Each input variable has a different degree of influence on the classifier's final labelling decision with respect to the case under consideration. Whilst F10, F6, and F12 have lower contributions to the classifier's decision, F5, F14, and F15 are identified as the major contributors resulting in the assignment and classification probabilities across the two classes. There is a 23.74% chance that perhaps C2 is the true label and the features responsible for this are the negative features, F15, F8, F13, F9, F11, F10, and F6. Driving the classifier's decision in favour of C1 are the positive features such as F5, F14, F1, F16, F7, F2, and F4.

{'C2': '23.74%', 'C1': '76.26%'}

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

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

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

Satisfied

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

SVC

C2

Health Care Services Satisfaction Prediction

The classification model employed made its label selection decision based on the information provided about the case under consideration. With a moderately low degree of confidence, it classifies the case under consideration as C2. Specifically, per the model, the probability of labelling the case as C1 is equal to 48.66%, hence not as likely as C2. The decision made here can be attributed to the influence of features such as F4, F1, F9, F15, and F2. However, F10, F7, F5, F11, and F16 are the least relevant features with respect to the classification made. The confidence level of the model is marginally above average and this can be attributed to the negative contributions of F14, F4, F3, F12, F10, F11, and F16. The negative features shift the prediction decision in the direction of C1, however, the positive contributions of other features such as F1, F9, F15, and F2 improve the odds of the C2 label.

{'C1': '48.66%', 'C2': '51.34%'}

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

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

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

Satisfied

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

SVC

C1

Flight Price-Range Classification

The prediction results are as follows: the probability that C1 is the correct label is 97.12%, the probability that C3 is the correct label is 2.55%, and the probability that C2 is the correct label is 0.33%. Judging based on the prediction probabilities across the classes, C1 is the most probable label. The very high confidence in the assigned label can be attributed to the very strong positive influence and contributions of the variables F7, F8, F4, F9, and F10. The other positive variables are F2, F11, and F3. The positive variables increase the probability that C1 is the correct label for the given case. Decreasing the probability of C1 are the negative variables F5, F12, F1, and F6. Considering that the combined effect of the negative factors is quite minimal in comparison to the top positive variables, it is not surprising that the model is very sure that neither C3 nor C1 is the best label for the given case.

{'C1': '97.12%', 'C3': '2.55%', 'C2': '0.33%'}

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

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

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

Low

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

DecisionTreeClassifier

C2

Insurance Churn

C2 is the model's predicted output for this given case, with an accuracy of 87.13% meaning the likelihood of C1 is only 12.87%. F13, F11, F15, F12, and F16 have the most effect on the output prediction choice in this case, whereas on the other hand, F4, F8, F6, and F7 are not that important to the decision made here. F13, F16, and F12 are the top negative features when you consider direction of their respective impacts, decreasing the model's reaction to labelling the given scenario as C2 and also F3, F14, F8, F6, and F7 are the other features that contribute negatively. In a nutshell, F11, F15, F2, F9, and F5 are primarily positive improving the odds of C2 with respect to this classification conclusion.

{'C2': '87.13%', 'C1': '12.87%'}

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

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

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

Stay

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

SVM_linear

C2

Employee Promotion Prediction

The classification model or algorithm classifies the provided data or case as C2 with a predicted likelihood of 94.16%, meaning that the chance of C1 being the true label is only 5.84%. The most relevant features driving the classification above are F9, F6, F4, F2, and F7, however, arranging the input features in-order of their contributions revealed that the least influential features are F3, F10, F1, and F5 since their values receive little consideration or emphasis from the algorithm. In relation to the directions of influence of input features, only F2 and F1 are shown to have negative contributions, which tends to drive the labelling judgement towards C1 instead of C2. Considering that the combined effect of all the negative features is lower than that of the positive features such as F9, F6, F4, F7, F11, and F8, it is valid to say that C2 is the most probable label.

{'C1': '5.84%', 'C2': '94.16%'}

{'F9': 'department', 'F6': 'avg_training_score', 'F4': 'KPIs_met >80%', 'F2': 'recruitment_channel', 'F7': 'region', 'F11': 'education', 'F8': 'length_of_service', 'F10': 'age', 'F1': 'no_of_trainings', 'F5': 'gender', 'F3': 'previous_year_rating'}

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

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

Promote

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

GradientBoostingClassifier

C1

Basketball Players Career Length Prediction

The final classification made was C1, but with a likelihood of only 55.19%, the model is uncertain about this prediction. By far, feature F6 had the most impact and following F6 are F16, F3, and F19 have been identified as having the comparable influence on classification. The combination of F6, F16, F3, F19, and F18 features has shifted the classification decision from C1 to C2. While F15, F5, and F14 are all features with a moderate impact on the classification, F15 is the only one of that set that has had a positive impact on the C1 classification and the remaining positives are F12, F17, and F11. Lastly, the features F13, F9, F1, F2, and F8 had very marginal negative contributions to the classification verdict.

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

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

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

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

Less than 5

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

BernoulliNB

C1

Personal Loan Modelling

Based on the prediction probabilities, C1 is the most likely label for the given case considering the values of the input variables and because the likelihood of C2 is very marginal, so the classifier is very confident that C1 is the right label. An analysis of the contributions of the variables has shown that F1 is the most relevant, with the strongest influence on the classifier's decision, however, to arrive at the classification above, the classifier probably ignores the values of the least ranked variables, F2 and F6. The level of confidence of the classifier with respect to the above classification decision is higher, primarily because most of the influential variables have a positive impact. F1, F5, and F8 are the top positive variables that increase the likelihood of C1. Having a different direction of influence, F3, F6, F2, and F7 are the negative factors, but compared to F1, their impact on the prediction decision above is low.

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

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

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

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

Reject

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

KNeighborsClassifier

C1

Real Estate Investment

Based on the information available about the case under consideration, the classification model is very uncertain about the appropriate labels for the case. According to the model, there is an almost equal distribution in terms of the probability that any one of C1 and C2 is an appropriate label. This indicates that any of the possible labels could be the true one, but for simiplicity, the model selects the class as C1. The above judgement is mainly due to the influence of the following factors or variables: F16, F12, F7, and F3 while the least relevant variables are F2, F11, and F17. Positive variables like F3, F7, F13, and F4 increase the model's response in favour of the assigned label. Nevertheless, negative variables such as F16, F18, F19, and F12 reduce the possibility that C1 is an appropriate label because their values support the selection of C2. Uncertainty about the classification here can be due to the fact that the most important negative properties, F16 and F12, have very high impacts, which moves the model's judgement away from C1 towards C2.

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

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

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

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

Ignore

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

LogisticRegression

C1

Tic-Tac-Toe Strategy

There is about an 81.01% chance that C1 is the probable label, hence the predicted probability for the C2 class is only 18.99%. The algorithm or classifier arrived at the prediction verdict above mainly based on the influence of features such as F7, F2, F1, and F4. For the algorithm, the least relevant feature is F9, which is shown to have a very small contribution in relation to the label choice here. When the directions of influence of the input features were investigated, it was discovered that F7, F6, F1, and F4 have positive attributions, pushing the algorithm higher towards the C1 label. Negative features such as F2, F3, and F5 assist in dragging or pushing the classification decision lower towards C1, where it was originally classified and this is mainly because their contributions to the prediction favour choosing or labelling the case as C2.

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

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

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

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

player B win

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

BernoulliNB

C1

Hotel Satisfaction

According to the classification algorithm, there is 77.69% chance that the given case is part of the C1 population. The features with the largest impact driving the algorithm to arrive at the above decision are F5, F2, and F6 which are followed in the decreasing order of influence by F3, F4, F9, F10, F15, F12, F1, F8, F7, F13, F14, and F11. Inspecting the direction of influence of the input features showed that, F5, F15, F12, F11, and F2 have negative influence on the prediction, shifting the algorithm's verdict towards the C2 class and can be blamed for the doubt in the classification decision. However, strongly pushing the classification higher towards the C1 label are the positive features such as F6, F3, F4, F9, F10, and F1.

{'C1': '77.69%', 'C2': '22.31%'}

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

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

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

dissatisfied

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

DecisionTreeClassifier

C2

Insurance Churn

The likelihood of the true label for the given test case being equal to the model's output prediction, C2, is 85.71% and since it's not 100%, there is a small chance of about 14.29% that the model could be wrong. Among the features employed for this classification, F9, F13, F1, F2, F14, and F7 are the top features influencing the model's prediction decision. The features with the strongest positive influence are F9 and F13 and in fact, these are shown to be the two main driving forces controlling the model's decision regarding the given case. Besides, some otf the other positive features include F2, F14, F3, F4, and F7. However, the atrribution of F1, F15, F8, F16, and F10 indicates the true label could perhaps be C1. While the different input features have some sort of contribution to the prediction made for this test case, the features F12, F6, and F11 have the least impact on the final decision here.

{'C1': '14.29%', 'C2': '85.71%'}

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

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

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

Leave

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

SVC

C2

Vehicle Insurance Claims

To begin with, the classification choice is entirely dependent on the information or data provided to the prediction model. According to the model, C2 has a 61.61 percent probability of being the true label, whereas C1 has a 38.39 percent chance of being the true label. Because the estimated probability of C2 is greater than that of C1, it is reasonable to assume that C2 is the most probable true label. The key variable responsible for this classification is F30, with a very significant positive effect on the model's conclusion, pushing it higher towards C2. F19, F20, F13, F10, F3, F7, F28, and F27 are the next set of relevant variables. F19, F13, F10, F7, F4, F8, and F28 have negative contributions that are responsible for the decrease in the chance that C2 is the actual label since they prefer to assign the C1 label instead. This means that the contributions of F20, F3, F17, F26, and F27, together with F30, can explain why the model is rather confident that C2 is the correct label.

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

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

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

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

Fraud

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

LGBMClassifier

C1

Employee Promotion Prediction

With a prediction likelihood of 62.34%, the model trained to generate predictions based on input variables identifies the presented example as C1. The model's label assignment choice for the given case is heavily impacted by the values of input variables such as F6, F1, and F4. The least important variables, on the other hand, are F2, F8, and F11. Furthermore, the impact of F10, F7, and F3 is regarded as moderate. F4 and F10 are the variables identified to have negative contributions to the classification when you take into consideration their respective direction of impact. All of the remaining variables have a positive influence, contributing to the classification of the presented case as C1. As a result, it is unexpected that the model's confidence is just 62.34% which suggest that the negative attributes may have a larger say in the appropriate label for the case under review.

{'C2': '37.66%', 'C1': '62.34%'}

{'F1': 'department', 'F6': 'avg_training_score', 'F4': 'recruitment_channel', 'F10': 'KPIs_met >80%', 'F7': 'no_of_trainings', 'F3': 'length_of_service', 'F9': 'age', 'F5': 'region', 'F2': 'education', 'F8': 'previous_year_rating', 'F11': 'gender'}

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

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

Promote

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

SVC

C2

Food Ordering Customer Churn Prediction

For the case under consideration, the model outputs C2 with high confidence level since the associated predicted class label is 89.73% whilst that of C1 is just 10.27%. Just few features out of the entire input features are shown to have control over the prediction made here. The prediction verdict C2 is mainly based on the variables F37, F5, F23, and F11. Other variables with moderate attributions include F38, F34, F3, F33, F46, and F31. Each variable mentioned above is shown to have different direction of contribution or impact for instance while F37, F11, F46, and F33 positively support the model's output decision, F5, F23, F38, F34, F31, and F3 contributed to decreasing the likelihood or odds of C2 being the true label for the given test instance. The variables shown to have no influence or contribution on the classification decision above are mainly F1, F28, F36, and F43.

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

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

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

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

Return

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

RandomForestClassifier

C1

Health Care Services Satisfaction Prediction

The model trained to solve the classification task labels the given case as C1 with a moderately high degree of confidence level equal to 60.13%. However, it is important to note that the prediction likelihood of C2 is 39.87%. Investigation of the contributions of the features to the above label assignment indicates that the most relevant features considered by the model are F1, F5, F14, and F12. Increasing the prediction likelihood of label C1 are mainly the positive features F1, F14, and F12. These features are termed positive features since their direction of influence is in support of the assigned label C1. On the contrary, F5, F7, and F3 are the top negative features, accounting for the uncertainty in the final prediction verdict. In plain terms, these negative features support labelling the case as C2, contradicting the model's decision in this case.

{'C1': '60.13%', 'C2': '39.87%'}

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

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

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

Dissatisfied

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

DNN

C2

Concrete Strength Classification

The following assertions are based on the information provided to the classification model. The classification model's confidence in this case's prediction output is approximately 69.40% and this suggest that the chance of label C1 is about 30.60%. The prediction attribution analysis shows that F1 and F6 are the most important features, whereas F2 and F4 are the least influential. F8, F5, and F3 are recognised as the only negative features considering the direction of effect of the features since their contributions reduce the prediction likelihood of the specified label, C2. F1, F6, F7, F2, and F4, on the other hand, have a positive impact on the model in favour of labelling the provided situation as C2 rather than C1.

{'C2': '69.40%', 'C1': '30.60%'}

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

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

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

Weak

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

GradientBoostingClassifier

C1

Broadband Sevice Signup

Because the chance that the label is the alternative class C2 is only 1.94 percent, the model anticipates that C1 will be the correct label in this situation. Specifically, it can be concluded that the model has a high level of confidence in the label C1. The feature attribution analysis conducted suggests that the two most relevant features considered when choosing the C1 are F15 and F32. F5, F40, F34, F38, and F22 were some of the other factors that positively helped with this prediction. F41, F21, F16, and F28, on the other hand, are the features with a negative influence on the above prediction judgement. In comparison to the F24, F22, F40, and F32, the foregoing features have little impact on the model and this might explain why the model is so certain that the correct label is C1. However, it is crucial to note that not all features are considered by the model during the label assignment with the irrelevant features such as F29, F3, F4, and F31 having extremely low attributions which happens to be almost zero.

{'C1': '98.06%', 'C2': '1.94%'}

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

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

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

No

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

RandomForestClassifier

C2

Student Job Placement

The model predicted that the example should be classified as C2 with a 76.06% likelihood but the model also identified that there was a 23.94% chance that the right label could actually be C1. The positive influence of features F9, F4, F3, and F12 on the model supports the class assignment of C2. Both F7 and F6 are features with a small positive impact on the classification decision for the given case. F1 and F2, in contrast, has a small negative impact on the output verdict that drives the decision away in favour of the other label. The features F8 and F11 have only a very small impact on the final classification decision. Finally, F10 is shown to have zero impact on the model in this case, hence it is not relevant to the prediction of class C2.

{'C2': '76.06%', 'C1': '23.94%'}

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

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

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

Not Placed

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

RandomForestClassifier

C2

Used Cars Price-Range Prediction

The classification model labels the given case as C2 at a very high confidence level since the probability that C1 is the correct label according to the model is only 3.50%. The assignment decision above is mainly based on the values of the features F3, F7, F8, and F5. On the other hand, the values of F9 and F6 are shown to have a very weak influence on the model's decision. The analysis revealed that only four of the input features support the decision by the model, while the remaining ones contradict the assigned label. The four positive features are F8, F5, F10, and F1.

{'C1': '3.50%', 'C2': '96.50%'}

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

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

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

High

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

SVM_linear

C2

Employee Promotion Prediction

The model generated the label, C2, with a very high likelihood of 99.69%, hence the probability that C1 is the right label is only 0.31%. Based on the analysis performed to understand the attributions of the different features, F2 was by far the most impactful positive feature whereas, the most negative feature is identified as F11. F10 also had a positive influence on the model's prediction, as did F5, F6, and F1. This is in contrast to F4 and F8, which had a negative influence on the prediction. Many of the features under consideration had only smaller impact on the outcome of the model and these are F7, F9, F1, F3, and F6. Considering the attributions of the input features, only F11, F4, F8, F7, F9, and F3 are shown to have negative attributions, decreasing the likelihood of the predicted label, however, the collective influence of the negative features is not enough to swing the model towards a different label.

{'C1': '0.31%', 'C2': '99.69%'}

{'F2': 'avg_training_score', 'F11': 'department', 'F10': 'KPIs_met >80%', 'F4': 'recruitment_channel', 'F8': 'age', 'F5': 'no_of_trainings', 'F7': 'previous_year_rating', 'F9': 'education', 'F1': 'region', 'F3': 'length_of_service', 'F6': 'gender'}

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

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

Promote

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

KNeighborsClassifier

C2

Advertisement Prediction

The item is labelled as C2 with a high degree of confidence since the predicted probability associated with the other class is 0.0%. Looking at the contributions of the features, only F6 and F7, are shown to drive the model towards predicting C1. However, these features are ranked as the least relevant, implying that their values have a very low impact on the model's decision. All the positive features, F4, F2, F5, F1, and F3, are ranked higher than the negative ones, with higher impacts on the model, significantly supporting the assigned label which could explain the high confidence level.

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

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

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

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

Skip

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

SVC_linear

C1

Personal Loan Modelling

With the prediction probability distribution across the labels, C2 and C1, respectively, equal to 0.30% and 99.70%, the model labels this instance as C1. The most important features are F1, F3, and F9. The variables, F4, F6, F2, and F5, have values, increasing the chances of C2 being the label for this case. Increasing the odds of C1 being the correct label are the values of the remaining variables. The strong positive variables are F1, F3, and F9 coupled with the moderate positive influence of F8 and F7 pushes the prediction in favour of C1 hence the prediction confidence level achieved.

{'C2': '0.30%', 'C1': '99.70%'}

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

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

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

Accept

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

DecisionTreeClassifier

C3

Car Acceptability Valuation

C3 is given as the predicted label with very high confidence, and according to the classification algorithm, there is no chance that either of the remaining three labels, C4, C3, and C2, is the right label for this case since the predicted probability of C1 is 100.0%. Based on the attribution analysis and investigations, the ranking of the input features from the most important to the least important is: F4, F2, F5, F6, F3, and F1. From the attribution analysis, F4 is the only one that positively contribute and support the above classification decision, while the remaining features such as F2, F5, F3, and F6 have negative contributions, shifting the decision in a different direction. In conclusion, looking at the predicted confidence level, one can say that the very strong attribution or influence of F4 is enough to dwarf the contributions of the features F2, F5, F6, F3, and F1.

{'C1': '100.00%', 'C4': '0.00%', 'C2': '0.0%', 'C3': '0.0%'}

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

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

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

Unacceptable

{'C1': 'Other B', 'C4': 'Acceptable', 'C2': 'Other A', 'C3': 'Unacceptable'}

KNeighborsClassifier

C2

German Credit Evaluation

In the present case, there is only a 12.50% chance that C1 is the correct label, which means there is an 87.50% chance that C2 is the true label. Therefore, the most probable class assigned by the model is C2. The above decision is mainly based on the influence of the following variables: F4, F6, and F7. Of these main variables, only F6 had a very strong positive impact on the model, increasing the prediction probability of the assigned label. The most important variables that lower the likelihood of C2 being the correct label are F7 and F4. The remaining two variables moving the decision away from C2 are F2 and F8. F1 and F9 are the least important variables, with a marginal impact on the model and this positive impact on the model is moderately low.

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

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

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

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

Good Credit

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

KNeighborsClassifier

C2

Tic-Tac-Toe Strategy

There is an evenly split chance that the prediction could be either of the two labels, C2 and C1. Based on the predicted probabilities, we can conclude that the model is uncertain about which label is the correct one. The abovementioned prediction decision is chiefly attributed to the influence of the following features: F7, F4, and F2, however, the least important or ranked ones are F3 and F8. The attributes F9, F6, F1, and F5 are shown to have moderate contributions.

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

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

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

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

player B lose

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

SVC

C1

Water Quality Classification

The label assigned to the given sample is C1 at a confidence level of 56.81%. This means that there is a 43.19% chance that the sample could be C2, representing an uncertain classification decision. The values of F7, F8, F4, F6, and F2 are the major contributing factors resulting in the classification decision here. On the other hand, the least important features are F9, F1, and F3, with a low level of influence. Considering the direction of influence of the features (that is, either supporting or contradicting the prediction above), only F4, F6, and F2 are shown to have positive attributions, increasing the likelihood of the assigned label. This implies that the values of the remaining features F5, F7, F8, F1, F9, and F3 have negative attributions, shifting the verdict in the opposite direction in favour of C2. In simple terms, the correct label should be C2 according to the negative features enumerated above.

{'C1': '56.81%', 'C2': '43.19%'}

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

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

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

Not Portable

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

SVC

C2

Australian Credit Approval

The classification algorithm classifies the given case as C2, since there is only an 18.57% chance that C1 is the correct label. The effects and contributions of positive input variables F8, F3, and F14 are the major drivers for the above classification. Besides, most of the remaining predictors such as F1, F13, F7, F2, and F9, are positive variables, decreasing the likelihood of the C1 label and making the label C2 more likely. The only variables with negative contributions are F10, F9, F5, and F4, which motivate generating the label C1 instead of C2. In summary, comparing negative attribution to positive attribution explains why the algorithm can determine that C2 is the right label for the given case.

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

{'F8': 'A8', 'F3': 'A9', 'F14': 'A14', 'F1': 'A12', 'F13': 'A7', 'F7': 'A4', 'F2': 'A5', 'F6': 'A11', 'F9': 'A1', 'F4': 'A13', 'F12': 'A10', 'F10': 'A2', 'F11': 'A6', 'F5': 'A3'}

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

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

Class 2

{'C1': 'Class 1', 'C2': 'Class 2'}

LogisticRegression

C1

Concrete Strength Classification

The odds are in favour of C1 being the correct label for the given case. This is because the probability of the other label, C2, is only 1.03%. Ranking the features in order of relevance to the classification decision above, F4, F2, F3, F1, F8, F5, F6, and F7. Among the set of features used for this prediction, F2, F8, and F5 are the only ones shown to decrease the likelihood of the C1 decision. The positive features increasing the chances of C1 being the correct label are F4, F3, F1, F6, and F7. The joint attribution of the positive features is stronger than that of the negative ones, which explains the confidence level associated with class C1.

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

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

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

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

Strong

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

GradientBoostingClassifier

C1

Paris House Classification

According to the prediction algorithm or model, there is almost 100% confidence that C1 is the label for the case under consideration. This is because the probability of C2 being the correct label is only 0.70%. The classification decision above is mainly based on the values of the following features: F3, F17, and F4 since their respective attributions are higher than any of the remaining features. F17 has a negative contribution to the prediction made by the model for this case, while in contrast, F3 and F4 have positive contributions, that push the classification decision in favour of C1. Unlike all the features mentioned above, the values of F13, F8, F14, and F1 have a limited impact on the classification decision above.

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

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

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

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

Basic

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

MLPClassifier

C2

Ethereum Fraud Detection

The C1 has a predicted probability of just 3.10 percent, but the C2 has a predicted probability of 96.90 percent, which implies that C2 is the most likely class chosen by the classifier for the supplied data. Not all of the input features are directly relevant to labelling the provided data and, per the attributions analysis, only F29, F27, F23, F8, F9, F38, F10, F1, F35, F36, F20, F14, F37, F12, F28, F32, F6, F13, F4, and F19 are the relevant features. However, F5, F17, and F15 are examples of irrelevant features since their contributions are mostly ignored by the classifier when classifying the given case. According to the attribution assessment, F29 and F27 have a very substantial combined positive influence, enhancing the classifier's response towards C2 rather than C1. In contrast, the top negative features are F23, F9, and F8, which weaken the classifier's response in favour of C1. When the attributions of F29, F38, and F27 are compared to the attributions of the negative features indicated above, it is not unexpected that the classifier is highly certain that C2 is the most likely label in this case.

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

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

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

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

Fraud

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

RandomForestClassifier

C2

German Credit Evaluation

The classification algorithm labels this instance as C2, but its level of confidence is moderate considering the fact that there is about a 44.0% chance that C1 could be the appropriate label. The features, F8, F3, F1, and F4, negatively influence the prediction verdict away from C2 and favour assigning C1 as the correct label. Contradicting the influence of the negative feature are features such as F2, F7, and F9, with positive contributions, improving the odds in favour of the probable label, C2. To summarise, the top features with the most influence on the above label assignment are F2 and F8, but F4 and F6 are the least influential input features considered by the algorithm.

{'C2': '56.00%', 'C1': '44.00%'}

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

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

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

Good Credit

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

GradientBoostingClassifier

C2

Food Ordering Customer Churn Prediction

The case given is labelled as C2 with close to an 82.07% confidence level, implying that the likelihood of C1 being the correct label is only 17.93%. The classification above is mainly due to the contributions of different features such as F11, F7, F42, F17, F41, and F18. But, not all features are considered by the classifier to arrive at the decision made for the given case. These irrelevant features include F44, F29, F12, and F25. Among the influential features as shown, F11, F7, F42, F17, and F32 are the top positives that increase the probability of C2 being the true label. However, F41, F18, F31, F3, F34, F24, F6, and F26 are the top negative features, driving the prediction lower towards C2 in favour of C1. In closing, the most important features with regard to this classification output are F11 and F7, all with positive attributions, explaining the very high confidence level.

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

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

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

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

Go Away

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

RandomForestClassifier

C1

Company Bankruptcy Prediction

The output labelling decision is C1 with almost 100% certainty, which indicates that there is practically no chance that C2 is the right label choice for the case under consideration. F34, F65, F86, F88, and F38 are the features with the highest joint positive impact, influencing the model's decision to output C1 and the feature F91 also has a high impact, but unlike F34, F65, F86, F88, and F38, F91 attempts to shift the decision away from C1 in the direction of C2. Also, F54 and F28 have a moderate impact on the decision towards C1, although this is still higher than features F59, F93, F47, and F8, which have a moderate impact, favouring the prediction of class C2. Besides, F28, F52, F48, F67, F55, and F68 all have a positive influence on the final classification verdict further increasing the odds in favour of the C1 label. It is worthy to note that for this classification decision, a large number of features are shown to be irrelevant hence received negligible consideration from the model, and these include F72, F77, F36, F49, and F51.

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

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

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

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

No

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

LogisticRegression

C1

House Price Classification

The label assigned by the classifier in this instance is C1, which had a very high prediction likelihood of about 99.93%. According to this classifier, the probability of C2 being the correct class is only 0.07%. Analysis performed shows that the confidence level of the classifier here is due to mainly the values of the features F9, F13, F8, and F1. The least relevant features to this classification verdict are F10, F5, F7, and F6 since the magnitude of their respective attribution is smaller compared to the remaining features. Furthermore, only the features, F2, F3, and F5, have a negative influence, increasing the chances of predicting the alternative label C2. However, when compared to the joint influence of the positive features such as F9, F13, and F8, the influence of the negative features is smaller, hence explaining the high degree of confidence in the predicted C1 label.

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

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

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

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

High

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

BernoulliNB

C2

Credit Card Fraud Classification

The classifier is very certain that C1 is not the accurate label for the given data or example, but that C2 fits. F4, F24, F16, F25, F14, F12, and F8 are the input features that have the most influence on the choice or judgment. F17, F29, F13, F20, F21, F26, F15, F27, F11, and F10, on the other hand, are found to be irrelevant and have negligible inlfuence on the classifier. Amongst the top features, F4, F24, and F16 are the one shown to have negative contributions, greatly favouring C1, lowering C2's prediction probability. Despite the significant negative attributions of the top impactful attributes, the classifier is quite certain that C2 is the correct label, based on the prediction probabilities.

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

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

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

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

Not Fraud

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

SGDClassifier

C2

House Price Classification

The classifier's anticipated label for this case is C2 which is a decision that it is highly confident about since the predicted likelihood is 100.0%. The most important variables are F3, F1, F10, and F6, whose values lead to the aforesaid classification conclusion. Under this classification instance, examination of the attributions of the features showed that F11, F8, and F2 are the least essential features. Because majority of the case's attributes positively validate the assigned label, it's not unexpected that the classifier picked the C2. F3, F1, F6, F12, F13, and F7 are all positive variables, while F10, F4, and F9 are three contradicting variables that moderately drive the labelling judgment towards C1.

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

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

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

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

Low

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

GaussianNB

C1

Tic-Tac-Toe Strategy

The model selects C1 as the correct label with a probability of 57.58%, while the other class, C2, has a slightly lower probability of 42.42%. The most relevant attribute is F4, followed by F3, F1, F8, F7, F6, F2, F5 and finally F9, which is the least relevant. The features F7, F2, and F4 have a positive influence, increasing the probability of the classification output, while F1 has a negative attribution, swinging the model to assign C2 instead. F8, F5, F3, and F6 are some of the other negative attributes. Finally, F9 has a very small positive control over the prediction in this test case but it further increases the confidence in the label chosen for the given case.

{'C1': '57.58%', 'C2': '42.42%'}

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

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

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

player B lose

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

SGDClassifier

C2

House Price Classification

The prediction verdict here is that the most probable class label is C2. Actually, the classification algorithm indicates that there is no possibility that the correct label is C1. Majorly contributing to the above classification are F8, F2, F3, and F4, all with positive influence. It is therefore not surprising that the algorithm is confident that C2 is the right label. The other positive features considered to arrive at the decision here are F12, F11, F13, F1, and F6. According to the attribution analysis, only F9, F7, and F5 have negative contributions, which tend to attempt to swing the final verdict in favour of C1. To sum up, the joint negative influence is not enough to outweigh the positive features, hence the C2 is assigned for the given case.

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

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

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

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

High

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

DecisionTreeClassifier

C2

Hotel Satisfaction

Due to the prediction probability distribution across the class labels, the labels assigned to this example is C2 with a high degree of confidence, close to 100 percent. The most significant features driving the classification above, according to the attributions of the input features, are F7, F9, F11, and F12. F14 and F10, on the other hand, are the least essential features to this prediction here. In addition, just four of the input features have a negative impact, skewing the classifier's judgement in favour of the C1 label. F14, F11, F13, and F10 are the opposing features. The contribution of the negative features, with the exception of F11, is quite modest when compared to the top positive features such as F9, F12, and F3.

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

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

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

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

satisfied

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

RandomForestClassifier

C2

Student Job Placement

The classification algorithm predicts that the data sample given should be classified as C2 with a probability of 76.06%, but it also finds that there is a 23.94% probability that the correct label will be C1. The positive influence of the F3, F1, F9, and F10 features on the algorithm supports the C2 class tasks. F5 and F7 are features with little positive influence on the classification decision for a particular case. F8 and F12, in contrast, has a small negative impact on the output decision that result in the reduction in the likelihood of C2 hence can be said to favour labelling the case as C1. F4 and F11 had only a minor positive impact on the final labelling decision and finally F6 was shown to have zero effect on the algorithm in this case.

{'C2': '76.06%', 'C1': '23.94%'}

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

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

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

Not Placed

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

KNeighborsClassifier

C2

Credit Risk Classification

According to the model, there is a higher chance that the case's label is C2. This prediction decision is based primarily on the attribution of the following features: F3, F1, F4, and F5. Aside from F5, all the other features listed above have a strong positive influence, increasing the probability of the predicted class C2. Similar to F5, the values of features F11, F2, and F7 suggest the other label, C1, could be the correct label. However, unlike F3, F1, and F4, each of the negative features has a moderate contribution to the final decision. The remaining features F10, F9, and F8 are shown to have marginal contributions to the model's decision for this case, and F6 was ranked as the least important feature. In summary, with strong positive attributions from F3, F1, F4, and F10, the model is very certain about the classification verdict, with a certainty of 100.0%.

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

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

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

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

Low

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

KNeighborsClassifier

C1

Company Bankruptcy Prediction

For the case under consideration, the model's output labelling decision is as follows: there is no possibility that C2 is the label for the given case, C1 is the most likely class label, with a confidence level close of 100.0%. The values of the input features, F87, F68, F67, F56, F50, F15, and F30, are the main driving forces resulting in the above classification. The features with moderate influence on the decision here are F10, F31, F62, F78, F52, F81, F92, F9, F93, F32, F80, F73, and F63. Apart from all the abovementioned input features, all the remaining ones, such as F65, F90, F27, and F17, are shown to be irrelevant to the decision made here. Also per the attribution analysis, not all the influential features support labelling the given case as C1, and these are referred to as negative features since they reduce the probability that C1 is the right label here and these are F30, F31, F80, F73, and F63. The notable positive features increasing the probability that C1 is the right label are F87, F68, F67, and F56.

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

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

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

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

No

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

KNeighborsClassifier

C1

Wine Quality Prediction

Based on the influence of features such as F3, F6, F7, and F5, the classifier is pretty confident that the correct label for the given data is C1, whilst, there is a 10.0% probability that the proper label could be C2. The majority of the features have positive contributions, while only F5, F10, and F11 are the negative features, decreasing the classifier's response towards choosing C1. The notal positive features that increase the classifier's response higher towards label C1 instead of C2 include F3, F6, F8, F9, F2, and F7. Taking into consideration the attributions of the input features, we can attribute the classifier's confidence associated with this prediction to the fact that the negative features only have a moderate impact on the classifier's decision for the given data.

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

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

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

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

high quality

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

LogisticRegression

C2

Music Concert Attendance

The model's prediction for this test case is C2 with an almost 100% confidence level which implies that the likelihood of it being a different class label is closer to 0%. Among the top influential feature-set, F15 has a value shifting the label choice in favour of C1, while the others, F18, F19, and F14, all have a positive impact supporting the decision made by the model to assign the label C2. Other features with positive support or impact on the prediction made include F9, F5, F12, and F20. However, F3, F10, F17, and F6 are the other negatives shifting the prediction decision in the direction of the alternative class label. TO sum up, the positive features clearly outweigh the negative features interms of their contributions, hence the confidence level in the classification output.

{'C2': '98.44%', 'C1': '1.56%'}

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

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

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

< 10k

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

MLPClassifier

C1

Ethereum Fraud Detection

The classification verdict for the selected case is C1, and the model is very certain about that considering the prediction probabilities across the possible classes. The top variables influencing this decision are F1, F7, F27, F29, and F13. Other variables that are regarded as somewhat important are F20, F32, F17, F26, F28, F10, F15, F36, F25, F38, F12, F24, F14, F35, and F2. Among the top variables, F1 and F7 decrease the prediction response; therefore, they are pushing the verdict toward C2. Similar to these features, F20, F32, and F28 negatively support assigning C1 to the case. Positively supporting the predicted label are the features F27, F29, F13, and F17. Unlike all the features mentioned above, the values of the remaining features such as F9, F4, F21, and F6, are unessential when determining the correct label for this case.

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

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

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

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

Not Fraud

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

LogisticRegression

C1

Employee Promotion Prediction

Classifying the given case based on the values of its features, C1 is the best label for the given case since its prediction probability is 99.45%, while C2's is just 0.55 percent. The most relevant factors for the classification or prediction declaration above are F4, F6, and F8, whereas the least influential factors are F7, F10, F3, and F1. The other factors' influence can be described as modest and after further inspecting the direction of effect of the factors, F4, F6, F9, F3, and F1 all contribute positively to giving the label C1. These are the favourable factors that raise the likelihood of C1 being the correct designation, however, F8, F2, and F5 are mostly responsible for minimising the chances of C1 and promoting C2.

{'C2': '0.55%', 'C1': '99.45%'}

{'F4': 'avg_training_score', 'F6': 'KPIs_met >80%', 'F8': 'department', 'F2': 'age', 'F9': 'no_of_trainings', 'F5': 'recruitment_channel', 'F11': 'previous_year_rating', 'F7': 'length_of_service', 'F10': 'education', 'F3': 'region', 'F1': 'gender'}

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

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

Promote

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

LogisticRegression

C1

Concrete Strength Classification

Per the predicted likelihoods across the classes, the model predicts label C1 in this case with a high confidence level. Features F6, F4, F2, and F7 are all driving the model towards the C1 classification, with feature F6 being the strongest driver and F7 being the weak driver among the above mentioned set of features. Features F8 and F5 have moderate negative impact on the C1 classification, while feature F1 has a strong positive impact. Finally, feature F3 has a very weak negative impact on the C1 classification decision driving the model towards assigning C2 to the case here.

{'C1': '90.65%', 'C2': '9.35%'}

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

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

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

Weak

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

KNeighborsClassifier

C1

E-Commerce Shipping

The classifier is very uncertain about the correct class for this example and this is because both classes are shown to be equally likely. The above prediction conclusion is mainly based on the influence of the top input features F3, F8, and F1, while F7, F6, and F2 have less influence on the classifier when classifying the given case. When the direction of influence or contribution of each input feature is examined, only F8, F8, F7, and F2 are revealed to have a positive contribution, improving the classifier's affinity to produce the label C1. The remaining features, F1, F3, F10, F4, F5, and F6 have a negative influence and contribution to the final decision.

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

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

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

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

On-time

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

LogisticRegression

C2

Airline Passenger Satisfaction

C2 is the label assigned to this data instance based on the fact that C1 is shown to be very unlikely, with a prediction probability of only 0.68%. The variables most relevant to increasing the probability of the prediction here are F1, F17, F19, and F6. Other positive features that increase the chances of predicting C2 are F20, F5, and F15, however, unlike F19, F17, F1, and F6, these have only moderate contributions to the model's classification decision for this instance. In contrast, F2 is the only top-ranked feature that led the model to classify towards C1, while other negative features with a moderately low contribution included F14, F11, F21, and F18. The least relevant features are F16, F4, F8, and F10, with a very low influence on the C2 prediction, however, unlike these features, F7 and F3 are shown to have no impact, since their attributions are very close to zero, when determining the correct label for the case under consideration. Finally, F3 and F7, according to the attribution analysis have no impact on the classification decision here.

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

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

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

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

neutral or dissatisfied

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

BernoulliNB

C2

Customer Churn Modelling

C2 is the class assigned to this case or instance. However, according to the classifier, there is a 5.75% chance that the other label, C1, is the correct one. The labelling decision above is mainly due to the values F4, F8, and F5. F3 and F10 are the least ranked features since they have marginal attributions. F8, F2, F7, and F4 have values, increasing the odds of C2 being the correct label and these four features are commonly known as positive variables given that they support the classifier's output decision for the given case. The remaining variables had negative attributions, driving the classification decision towards label C1 and the most negative variables are F5, F1, and F6.

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

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

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

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

Stay

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

LogisticRegression

C1

Real Estate Investment

The model predicts the class label of this test case or instance as C1 and it is quite confident in the above prediction decision considering the predicted confidence level. The above prediction decision was made primarily based on the values of the following features: F9, F19, F15, and F14. The top features, F9 and F19, positively contribute to the final prediction of C1. Besides, F14 also has a positive impact, pushing the model to output C1. However, the value of F15 supports the prediction of the alternative label, C2. However, compared to F9 and F19, the influence of F15 is very small. The features with moderate influence or impact on the prediction made for this test case are F2, F8, and F1. While F2 moderately supports the C1 prediction, F8 and F1 have values, pushing the model toward predicting C2.

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

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

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

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

Invest

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

DecisionTreeClassifier

C2

Concrete Strength Classification

The case is labelled as C2 by the classification model, and according to the model, there is little to no chance that the correct label could be C1. Per the feature attribution inspection, F2 and F1 are the least influential features. The classification decision to label this case as C2 is mainly due to the positive contributions of F4, F3, and F7. However, the strong negative influence of F8 indicates that the true label could be C1, but since the likelihood of C1 is 0.0%, we can say that the positive features successfully drive the decision in favour of the C2 label. F6, F5, and F1 are the other negative features that unsuccessfully attempt to shift the decision in favour of C1. From the attribution analysis and the predicted likelihoods across the classes, we can conclude that the model is certain that C1 is not the true label.

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

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

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

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

Strong

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

SVC

C1

Real Estate Investment

The decision of the classification model on the true label with respect to the given case is based on the information provided to it. From the prediction probabilities, C1 is selected by the model as the most likely label, with a very high confidence level equal to 97.49%. According to the attributions analysis, the very high confidence in the validity of C1 can be attributed to the very strong positive influence of F2, F9, and F8. The contributions of all the other features are moderate to low. The least relevant features are F15, F16, F11, and F5, whereas the moderate ones include F1, F3, F13, and F7. The very marginal uncertainty with respect to the classification decision here can be blamed on the moderate influence of negative features such as F1, F3, F13, F20, F6, and F7. Aside from F2, F9, and F8, some of the other positive features are F14, F19, and F16, with moderate to low contributions, pushing the decision further higher towards C1 away from C2. Finally, F5 has a negligible contribution to the decision above.

{'C2': '2.51%', 'C1': '97.49%'}

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

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

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

Invest

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

KNeighborsClassifier

C3

Cab Surge Pricing System

With a moderate likelihood of 50.0%, the label for this case is judged to be C3. The classifier, on the other hand, says that C1 and C2 are equally likely, with a predicted probability of 25.0 percent. The aforementioned decision is mostly dependent on the features of the given case and the values of F4, F7, and F10 are demonstrated to be the primary factors influencing the classification output decision. When compared to F4, F7, and F10, the other variables, such as F9, F8, and F5, have lower attributions. According to the attribution assessment, F4, F7, F10, F8, and F11 are the factors that positively contribute to the choice, implying that they are the ones that push the classification closer towards C3. F9, F5, F3, F1, and F2, on the other hand, are the top negative factors that sway the choice somewhat toward the other labels, C1 and C2. In fact, it is because of these negative variables that the classifier presents the probabilities across the C2 and C1.

{'C1': '25.00%', 'C2': '25.00%', 'C3': '50.00%'}

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

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

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

C3

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

LogisticRegression

C2

Music Concert Attendance

With a prediction probability of around 82.06 percent, the algorithm predicts class C2. In the aforementioned prediction judgment, F4, F11, F13, and F3 are all important. The top positively contributing features supporting the C2 prediction are F4, F11, and F3, while F13 is pushing the final prediction away. F15 also has a positive impact on the categorization, but F7 has a negative impact and finally, F6, F9, F12, and F17 have very little influence on the algorithm among the features, when picking the most appropriate label in this case.

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

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

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

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

> 10k

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

LogisticRegression

C2

House Price Classification

For this test case, the model predicts C2 with 99.93% certainty and what this means is that there is only 0.07% chance that C1 could be the right one. The features with the highest impact are F4, F13, F2, and F9, which are all shown to contribute positively to the prediction decision mentioned above. While F6 and F8 support the prediction, F3 is the feature with the strongest negative support for the prediction. Of the features with a small impact, namely F11, F7, F1, F12, F10, and F5, only F7 and F12 negatively support the prediction while the others positively support it.

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

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

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

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

High

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

BernoulliNB

C1

Employee Promotion Prediction

This model trained on eleven attributes predicts class label C1 for this case with a confidence level equal to 54.21%. This suggests that the likelihood of C2 being the correct label is 45.79%. The classification decision above is mainly based on the influence of the features F9, F5, F1, and F6. The most relevant features are the negative features, F9, F5, and F1. These features are regarded as negative features given that their values are shifting the prediction decision in the direction of C2. The positive attributes are F6, F4, F11, F3, and F8, supporting the model's prediction for this case.

{'C1': '54.21%', 'C2': '45.79%'}

{'F9': 'KPIs_met >80%', 'F5': 'previous_year_rating', 'F1': 'avg_training_score', 'F6': 'department', 'F4': 'education', 'F11': 'recruitment_channel', 'F2': 'no_of_trainings', 'F7': 'length_of_service', 'F3': 'region', 'F10': 'age', 'F8': 'gender'}

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

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

Ignore

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

LogisticRegression

C2

Used Cars Price-Range Prediction

With a moderate confidence level of 67.95%, the model predicts C2 for the case under consideration, but it is important to consider the fact that there is a 32.05% chance that C1 could be the correct label instead. The most influential variables resulting in the aforementioned classification decision are F6, F9, and F3. While F6 and F9 have negative contributions towards the C2 prediction; favouring the assignment of C1 instead, F3 is the top positive contributing feature. F5, F1, and F2 had a small positive effect on prediction, whereas F7 had a smaller negative effect. Finally, F4 is the least relevant variable, and therefore, its negative attribution has no significant influence on the model with respect to the given case.

{'C1': '32.05%', 'C2': '67.95%'}

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

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

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

High

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

LogisticRegression

C2

Employee Promotion Prediction

As per the classification algorithm, the most appropriate label for the given case is C2 because its prediction likelihood is 99.45%, whereas that of C1 is only 0.55%. For the classification or prediction assertion above, the most important variables are F2, F1, and F9, while the least influential variables are F5, F7, F3, and F11. Regarding the direction of influence of the variables, the ones with positive contributions to assigning label C2 are F2, F1, F8, F3, and F11 which in fact increase the odds of C2 being the correct label. Finally, decreasing the odds of C2 and supporting C1 are mainly the values of the variables F9, F6, and F4.

{'C1': '0.55%', 'C2': '99.45%'}

{'F2': 'avg_training_score', 'F1': 'KPIs_met >80%', 'F9': 'department', 'F6': 'age', 'F8': 'no_of_trainings', 'F4': 'recruitment_channel', 'F10': 'previous_year_rating', 'F5': 'length_of_service', 'F7': 'education', 'F3': 'region', 'F11': 'gender'}

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

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

Promote

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

SGDClassifier

C1

House Price Classification

C1 is the label predicted by the classification model employed and looking at the prediction probabilities, it valid to concluded that the model is very certain about the selected label. The features considered most relevant by the model for the above decision are F1, F10, F7, and F6, while those with the least consideration are F3, F4, and F13. On the basis of the analysis, majority of the input features positively affirm the prediction for this case; therefore, it is not surprising that the model chose the C1 label and the positive features include F1, F7, F6, F11, F2, F8, and F5. The three negative features that moderately bias the labelling decision towards C2 are F12, F10, and F9.

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

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

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

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

Low

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

LogisticRegression

C1

E-Commerce Shipping

The confidence level for the prediction made for the given case is 71.57%. F4 has a significant impact on the outcome in the negative. The values F8, F7, F6, F2, F3, F5, and F10 all have a positive impact on the results, but they are still less than the effects of F4. The analysis shows that F4 has the highest impact on the model's prediction decision here, it has an overwhelmingly negative effect. F7, F6, F2, and F3 have a positive effect on the model's prediction. Because of the strength of the F4 feature, all other features have little effect on the outcome. In addition, the uncertainty in the prediction could be attributed to the pull of F4, which drives the model to predict an alternative label.

{'C1': '71.57%', 'C2': '28.43%'}

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

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

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

On-time

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

LogisticRegression

C1

E-Commerce Shipping

53.78% and 46.22%, respectively, are the chance or likelihood of any of the classes C1, and C2 being the appropriate label for the case given here. As a result, it's safe to say that C1 is the most likely label for this situation and F4 is identified as the most influential feature whereas F6, F8, and F10 have very low contributions to the decision made by the classification algorithm with respect to the given case. In addition, F5, F7, F3, F1, F9, and F2 have moderate contributions higher than F6, F8, and F10 but lower than F4. Despite the strong positive influence of F4 and F7 supporting the assignment of C1, the negative influence of F5, F3, F1, F2, and F10 shift the classification judgment fairly towards the C2 label which explains the 46.22% likelihood.

{'C2': '46.22%', 'C1': '53.78%'}

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

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

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

Late

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

LogisticRegression

C2

Used Cars Price-Range Prediction

The output decision for the provided data is C2, with a very high confidence level, based on the output prediction probabilities across the two classes since C1 has a probability of around 0.00%. F5, F8, and F1 are the most influential factors in the above-mentioned label assignment, however F7 and F9 are the least influential. The unusually high degree of confidence associated with the classification choice in this case might be attributable to the fact that the bulk of the input variables exhibit attributions that improve the model's responsiveness towards label C2. F4, F10, and F7 have only the negative contributions, attempting to persuade the model to classify this case as C1. To cut a long story short, the joint contribution of the negative variables is quite low in comparison to that of the positive variables, resulting in the model's certainty in the decision above.

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

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

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

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

Low

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

SVC

C1

Advertisement Prediction

When given the task of labelling the given case one of the possible labels, C1 and C2, the model assigns C1 as the most likely correct label, with a confidence level of roughly 99.90%. This degree of confidence indicates that the likelihood of C2 being the right designation is merely 0.10%. According to the attribution analysis, each variable has a distinct degree of effect or contribution to the model's arriving at the above-mentioned classification. F7, F1, F5, and F4 are the features accounting for the model's extremely high confidence in the assigned label. In fact, the only input variables having a negative impact are also the least relevant ones, F3 and F6.

{'C1': '99.90%', 'C2': '0.10%'}

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

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

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

Skip

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

LogisticRegression

C1

Concrete Strength Classification

According to the classification model employed here, the most probable label for the given case is C1 with a confidence level equal to 98.97%. Per the attributions analysis, F8 and F7 are the most significant and influential features driving label selection. The least ranked features are F2 and F6, while F4, F5, F1, and F3 have moderate contributions. Negatively supporting the above classification output are F7, F1, and F3, pushing the model to assign the alternative label. However, given the fact that the prediction probability of C2 is only 1.03%, it can be concluded that the joint positive influence of F8, F4, F5, F2, and F6 strongly drives the model to label the case as C1 instead of C2.

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

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

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

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

Strong

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

KNeighborsClassifier

C2

Wine Quality Prediction

The classifier is quite sure that the right label for the data given is C2 based on the influence of variables such as F9, F7, F5, and F4. There is a 10.0% chance that the correct label is C1 and per the attributions examination conducted, the bulk of the traits contribute positively, with only three contributing negatively. The negative variables are F4, F6, and F3, which reduce the classifier's preference for C2. F9, F7, and F5 are notable positive variables that boost the classifier's response to outputting C2 rather than C1. All in all, the classifier's confidence in this prediction may be attributed to the fact that the negative variables only have a minor influence on the prediction choice here.

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

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

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

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

high quality

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

DecisionTreeClassifier

C2

Vehicle Insurance Claims

C2 was assigned to the given case by the classifier with a likelihood of 93.32%, leaving thhe likelihood of the C1 equal to only 6.68%. The most influential features were F15, F25, and F21. The remaining features with non-zero attributions are F5, F31, F18, F17, F7, F9, F14, F19, F6, F33, F23, F27, F1, F26, F29, and finally F2. F15 and F25 were highly influential in the positive direction, increasing the odds of the predicted label being correct, whereas F21 had a negative impact, driving the prediction in favour of a different label. Furthermore, F5 had a positive impact on the prediction, whereas F31 and F18 negatively influenced the prediction. Finally, the features that we can say have no impact at all on the prediction made here are as follows: F32, F3, F8, F12, and F10.

{'C2': '93.32%', 'C1': '6.68%'}

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

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

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

Not Fraud

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

AdaBoostClassifier

C2

Basketball Players Career Length Prediction

With moderately high confidence, the classifier indicates that the most probable label for the given data is C2 with only just a 21.80% chance that it could be C1. The main driving features for the above classification or prediction decision are F19 and F11. The remaining features such as F8, F12, F14, and F4 have moderate to low influence on the above decision. Inspecting the attributions of the the input features showed that the ones with negative impact or contribution are F8, F4, F13, F2, and F6. From the attributions, we can see that the remaining features have positive contributions or influence and as a matter of fact, the certainty of the classifier for this classification can be attributed mainly to the strong positive contributions of F19 and F11 coupled with the contributions of the other positive features such as F12, F14, F1, and F15.

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

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

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

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

More than 5

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

GradientBoostingClassifier

C1

Health Care Services Satisfaction Prediction

Given the fact that the likelihood of C2 being the correct label for the case under consideration is only 36.34%, the model assigns the label C1. The prediction decision between the two classes is highly based on the values of the features F3, F11, F6, and F2, whereas those with the least attributions or contributions regarding this label assignment are F8 and F16. Among the top influential features, F3 and F11 have very strong positive contributions, increasing the probability of the label C1, while the value of F6 value suggests the other label, C2, could be the true label. This pull or shift towards label C2 is further supported by the values of F14, F4, F13, F10, F8, and F5. Conversely, the remaining features, together with F3 and F11, positively encourage the prediction of C1.

{'C1': '63.66%', 'C2': '36.34%'}

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

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

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

Dissatisfied

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

KNeighborsClassifier

C2

Cab Surge Pricing System

C2, out of the three potential classes, is the the label assigned with a high probability of 50.0%. However, the classifier indicates that C1 and C3 are equally likely, with a predicted probability of 25.0%. The aforementioned judgement is mostly based on the variables of the given case. The variables F1, F4, and F2 are shown to be the main factors resulting in the classification output decision. The remaining variables, such as F11, F8, and F9, have lower attributions compared to F1, F4, and F2. The attribution analysis also indicated that F1, F4, F2, F8, and F7 are the variables that positively contribute to the decision, meaning they are the ones that shift the classification higher towards C2. On the contrary, F11, F9, F12, F3, and F5 are the top negative variables that steer the decision slightly towards the other labels, C1 and C3. In fact, it is because of these negative variables that the classifier indicates the probabilities across the C3 and C1.

{'C1': '25.00%', 'C3': '25.00%', 'C2': '50.00%'}

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

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

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

C3

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

LogisticRegression

C2

Flight Price-Range Classification

The chances of selecting the correct label from one of the possible labels C1, C3, and C2 are 18.51%, 5.86%, and 75.63%, respectively. As a result, it can be deduced that the classifier's anticipated label in this situation is C2. The values of the input features were used as the basis to make the aforementioned prediction judgments. Some of these features have values that positively support the assigned label, while others have values that contradict the classifier's decision, driving it toward one of the other two labels. F1 is the most influential feature, following which are the variables F4, F10, F5, and F2, enumerated according to their respective relevance to the aforementioned label selection. F1, F5, and F2 are positive features that increase the classifier's response towards generating the C2 label, but F4 and F10 are negative features, lowering the odds of C2 being the correct label. F9, F7, F3, F8, and F12 are features that have a moderate influence on the classifier in this case, while F6 and F11 have only a marginal impact. F4, F10, F3, and F11 are the features that have values supporting the assignment of any of the other labels, while the rest favour the C2 prediction, therefore, the predicted probabilities across labels is unsurprising. Furthermore, the predicted likelihood of C2 is higher than all the other labels which is attributed to the fact that the positive features' combined impact is bigger than negative features' combined impact.

{'C3': '5.86%', 'C1': '18.51%', 'C2': '75.63%'}

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

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

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

High

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

GradientBoostingClassifier

C1

Broadband Sevice Signup

In this case, the model expects C1 to be a label since the probability that the label is the alternative class C2 is only 1.94%. This means that the model has a lot of confidence in the selected label, C1. F30 and F23 are the two most important prediction variables positively controlling the assignment of C1 in this case. Other variables that contributed positively to this prediction included F15, F25, F32, F40, and F20. On the other hand, the values F42, F24, F37, and F8 constitute a feature set with a negative impact on the above prediction decision. However, the above features have little effect on the model compared to the F1, F20, F25, and F23, which may explain why the model is confident that the true label is probably C1. Finally, for the case under consideration, F18, F9, F29, F22, F41, and F33 are some of the features, with practically no effect on the prediction decisions of the model, hence they can be considered negligible to the classification here.

{'C1': '98.06%', 'C2': '1.94%'}

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

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

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

No

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

KNeighborsClassifier

C1

Credit Risk Classification

The following classification assertions are based on the information provided on the case under consideration. The most probable or likely label judged by the classifier is C1 since its prediction probability is 60.0% compared to the 40.0% of C2. The influence of the features on the classifier's decision here can be ranked in the order F3, F10, F4, F9, F8, F7, F1, F6, F2, F5, F11. In fact, with the exception of F11, all the features are shown to have attributions, resulting in the predicted probabilities across the labels. The F3, F10, F4, and F2 have negative contributions, leading to the classifier's confidence in the validity of the C1 label and this is because they are the features that support labelling the case as C2. However, the positive features F9, F8, F7, F1, F6, and F5 tip the scales higher in favour of C1. Since the most influential features F3, F10, and F4 have negative contributions, it is not surprising that the classifier has the probability of C2 equal to just about 40.0%.

{'C2': '40.00%', 'C1': '60.00%'}

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

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

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

High

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

SGDClassifier

C1

Airline Passenger Satisfaction

At a confidence level of 100.0%, the model labels this case as C1 and what this indicate is that there is no chance for C2 to be the correct label given the values of the input features. The above classification decision can be attributed to values for features such as F16, F15, F6, F19, F17, and F22. For this C1 prediction, the most important features are F16, F15, and F6. These are all positive features, meaning they strongly support the model's decision with respect to the case under consideration and a further push towards the assigned label is offered by the contributions of the other positive features such as F17, F22, F18, and F8. On the other hand, shifting the decision in the opposite direction are the negative features such as F19, F10, F14, F3, and F12. However, compared to F16, F15, and F6, the joint influence of the negative features mentioned above is weak. Finally, the values of the features F4 and F13, both with almost zero attributions, are not relevant when it comes to deciding the correct label for this case.

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

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

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

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

satisfied

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

SVC

C2

German Credit Evaluation

For the case under consideration here, there is a 70.83% probability that the true label is C2 and what this means is that there is also a 29.71% chance that C1 could be the correct label. Among the features, the top two most impactful are F8 and F9. The next features, ranked in order of the magnitude of their respective attribution are F1, F7, F3, F6, F4, F2, and F5. Out of the nine features, only three of them have values pushing for the prediction of label C1 while the rest are referred to as positive features given that their values motivate the prediction of class C2. The three attributes with the negative impact, shifting the prediction decision away from C2, are F9, F1, and F7. The collective influence of positive features is higher than that of negative features F9, F1, and F7.

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

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

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

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

Good Credit

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

BernoulliNB

C1

German Credit Evaluation

The algorithm labels the data given as C1 and the prediction probabilities across the possible labels C1 and C2, respectively, are 51.39% and 48.61%. Judging based on the prediction probabilities, the algorithm shows signs of uncertainty in the above decision. F4, F6, F9, and F3 are the primary contributors to the classification verdict here. The contributions of F3, F7, and F1 are moderate, while those of F8, F2, and F6 are lower compared to the other variables. Positively supporting the classification are F4, F9, F2, and F5, while all the remaining variables have a negative impact that decreases the probability of C1 being the correct label. F6, F3, and F7 are negative variables that can be blamed for the uncertainty in the classification decision being made here.

{'C1': '51.39%', 'C2': '48.61%'}

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

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

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

Good Credit

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

GradientBoostingClassifier

C2

Health Care Services Satisfaction Prediction

The label assignment decision is solely based on the values of the different input features passed to the classification algorithm since the values of these features are used as the basis to make the prediction judgments. The likelihood of any of the classes C2 and C1 being the correct label is 76.26% and 23.74%, respectively, therefore, it is valid to assert that the true label for this case is C2. From the attribution analysis, F15, F11, and F12 have the highest contribution to the decision, whilst F4 and F8 are the least relevant features. In between these two ends are the moderately influential features, such as F5, F9, F14, F7, and F10. Furthermore, the negative features F11, F14, F3, F6, F2, F1, and F4 can be blamed for the fact that the algorithm is not 100.0% certain about the labelling decision and this mainly because the negative features contribute towards choosing C1 instead of C2. Conversely, the positive features such as F15, F12, F5, F9, F7, F10, and F13 are the ones driving the decision higher towards C2.

{'C1': '23.74%', 'C2': '76.26%'}

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

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

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

Satisfied

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

LogisticRegression

C2

Broadband Sevice Signup

Here the classifier labels the given case as C2 with a moderately high confidence level. Specifically, the prediction likelihood of class C1 is only 21.67%. The main drivers for the classification above are F23, F25, F3, and F7. Among these top features, F23 and F25 have the most significant influence on the classification outcome, and they happen to have positive contributions, increasing the likelihood of class C2. On the other hand, the F7, F3, and F24 have a moderate negative contribution, reducing the odds of a C2 prediction. F28, F10, F29, and F26 are other notable positive features, while F32, F8, F19, and F21 are notable negative features. However, the classifier did not take into account all of the input features when arriving at the above-mentioned classification verdict; the features including F37, F11, and F18 are deemed irrelevant. To summarise, considering the attributions of influential features such as F23, F25, and F7, it is evident why the classifier is quite certain that C2 is the most probable label for the given case.

{'C1': '21.67%', 'C2': '78.33%'}

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

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

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

Yes

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

LogisticRegression

C3

Cab Surge Pricing System

The label assigned in this case by the classifier is C3, with a moderately high prediction confidence of 66.11%. Since the confidence level with respect to this C3 is not 100.0%, it is possible that one of the other labels is the true or correct label, and C2 is the next most likely label. The input variables F2, F3, F7, and F4 have a significant impact on the abovementioned prediction judgement. The value of features F2, F7, F1, and F11 contributes positively to the C3 label, instead of the other labels. F3, F4, F10, and F6 are the variables having a contradictory influence, shifting the final decision in the direction of the other labels. The remaining positive variables are F8, F9, F12, and F5. Of all the predictors, the ones that contributed the least to the prediction included F10, F12, F6, and F5. In summary, given the attributions of the predictors, it is clear why the classifier indicates that C3 is the correct class in this scenario.

{'C2': '31.78%', 'C3': '66.11%', 'C1': '2.11%'}

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

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

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

C2

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

SVM_poly

C4

Mobile Price-Range Classification

The classification assertions arrived here are mainly based on the influence and contributions of the different input variables. The prediction probabilities across the four possible classes C1, C2, C3, and C4 are 0.05%, 0.04%, 0.47%, and 99.45%, respectively. Therefore according to the classifier, the most likely class label for the case under investigation is C4 and it is quite sure that neither C3 nor C1 nor C2 is the true label here. The influence of F17 is shown to be the major contributing factor resulting in the prediction decision made by the classifier and the contributions of the remaining features such as F6, F13, F7, and F3 are moderately low compared to that of F17. The strong positive influence of F17 coupled with other positive features such as F7, F8, and F3 can explain the very high confidence level in the prediction decision. On the flip-side, the input features F6, F13, and F15 are considered negatives since their attributions marginally reduce the prediction probability of the C4 label.

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

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

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

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

r1

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

BernoulliNB

C2

Job Change of Data Scientists

The prediction likelihood of class C2 is 84.87%, making it the most probable label for the given case. When making the above prediction, the most relevant features considered are F5, F7, F3, and F8. Conversely, F11, F6, and F9 are the least influential features, with their values receiving little consideration from the model regarding this classification. Assessing the direction of influence or contribution of the features suggest that there is a split between the number of features with a negative influence and those with a positive influence. However, only two of the negative features, F7 and F1, have a somewhat high influence; the others , F12, F11, F6, and F2, have a lower negative influence. To put it concisely, the combined influence of the positive features, such as F5, F8, F4, F10, and F3, outweighs that of all the negative features combined, therefore, it is entirely plausible to see such confidence level of the model for the classification here.

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

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

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

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

Leave

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

MLPClassifier

C2

Annual Income Earnings

With respect to the given case, the most probable label for the given case is C2, with a 99.81% chance of being the correct label, therefore the probability of C1 is only 0.19% for this case. Among the input variables, only four features are shown to have a negative influence on the classification decision above: F2, F8, F5, and F6 since their contributions to the decision only favour labelling the given case as C1 instead. On the flip side, pushing the classification strongly towards C2 are the features F13, F4, F10, and F1 explaining the very high confidence in the choice of label assigned here.

{'C2': '99.81%', 'C1': '0.19%'}

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

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

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

Under 50K

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

GaussianNB

C1

Tic-Tac-Toe Strategy

For a particular case, the model predicted the class designation C1 with 75.50% confidence. Based on the attributions analysis, the feature that had the biggest impact on the final labelling decision were the F1 and F3, which happened to strongly support the assignment of label C1. Contributing differently to F1, the feature F5 is the top negative feature, reducing the odds that C1 is the correct label. F4, F2, F5, and F7 have similar influences on the model in terms of the magnitude of their contributions or attributions, however, the directions of their respective effects are different: the features F4 and F2 positively support the model, driving the prediction towards class C1, while F5 and F7 work against it. F8, F9, and F6 are features that have little effect on the model when assigning the label for the given case, and all of them negatively contributed to the C1 class selection. Among all the features with little contribution to the prediction verdict above, F6 is the least relevant.

{'C2': '24.50%', 'C1': '75.50%'}

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

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

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

player B win

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

RandomForestClassifier

C4

Mobile Price-Range Classification

The model reveals that C2 and C1 each has a zero prediction probability, while C3 has a 3.85%. This indicates that C4 is the most likely label for the present context with approximately 96.15% certainty. F2, F9, and F7 are the most important elements driving the above classification, whereas F10, F16, F11, F14, and F5 are the least important. The intermediate elements, which comprise F8, F6, and F1, have varied degrees of influence, ranging from moderate to low. F8 is the only with a negative contribution among the top influential features, F2, F9, F7, F8, and F6, skewing the forecast slightly towards a different possible label. Furthermore, the top two positive elements, F9 and F2, have a greater effect than the sum of all the negative ones.

{'C2': '0.00%', 'C1': '0.00%', 'C3': '3.85%', 'C4': '96.15%'}

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

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

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

r4

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

LogisticRegression

C1

House Price Classification

The prediction is that class label C1 is very likely the correct label, given that the associated confidence level is 99.93%. The features F11, F7, and F3 appear to have very smaller or little impact on the prediction of C1 compared to F5, F12, F4, F10, and F8, according to the attribution analysis. F5 and F12 are the features with the highest impact on the model's output prediction verdict above and fortunately the values of these features positively support the C1 classification verdict. Other positive features increasing the odds in favour of C1 include F4, F10, F1, and F2. On the contrarily, the feature F8 negatively influences the model's prediction of C1, shifting the verdict in the opposite direction. It is important to note that, only the features F8, F13, and F3 have negative attributions, while all the remaining ones have positive attributions. The joint positive attribution outweighs the negative attributions from F8, F13, and F3.

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

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

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

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

High

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

GaussianNB

C1

Tic-Tac-Toe Strategy

For the given case, the model predicted the class label C1 with a certainty of around 75.50%. By far, the feature with the most impact on the final classification was F3, which positively supports the decision. Feature F1 was the feature that contributed the most to pushing away the classification decision from C1, that is, they are decreasing the likelihood of C1 being the correct label. F2, F8, F1, and F9 all had a similar impact on the classification. However, the direction of influence is different, with features F2 and F8 pushing the model's decision to class C1 and features F1 and F9 doing the opposite. F6, F5, and F7 are the features that had closer to negligible impact on the final classification, all of which had a negative contribution towards class C1.

{'C2': '24.50%', 'C1': '75.50%'}

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

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

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

player B win

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

LogisticRegression

C1

Food Ordering Customer Churn Prediction

Judging based on the values of the input features, a decision is made by the classifier to label the given data as C1 with a prediction confidence equal to 84.90%. The major influential features resulting in the classification here are F19, F15, F35, and F38. F19 and F15 are identified as the most negative features, with contributions that lead to a decrease in the classification confidence of label C1. F38 and F35, on the other hand, are the top positive features, leading the classifier to label the case as C1. Other notable negative features are F28, F42, and F39 while other notable positives are F6, F45, F5, and F24. Unlike all those mentioned above, F11, F36, F8, and F1 are among the many irrelevant features with negligible contributions to the classification decision here.

{'C1': '84.90%', 'C2': '15.10%'}

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

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

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

Return

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

LogisticRegression

C2

Used Cars Price-Range Prediction

The prediction model predicts C2 for the case under consideration since the likelihood of C1 which is equal to 30.05%, is lower than that of C2 and this verdict came about mainly based on the values of the input features passed to the model. F3, F1, and F7 are identified as the most influential features with higher impact on the model's labelling decision here and among them F3 and F1 have negative contributions decreasing the model's response towards the assigned label. Furthermore, F7, F8, and F10 have a positive impact on the model and in effect pushes the decision higher towards C2, while F5, F6, and F4 have identical direction of impact as that of F1 and F3. Finally, F2 is the least relevant feature, therefore, its negative attribution has little effect on the model in this case and also the positive influence of F9 further supports the assigned label.

{'C1': '30.05%', 'C2': '69.95%'}

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

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

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

High

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

DecisionTreeClassifier

C2

Credit Risk Classification

The model assigned the label C2 to the given instance since its associated likelihood is far higher than C1. The most relevant features controlling the prediction decision above are F9, F5, and F10. The less relevant ones include F3, F4, and F1. The majority of the features have values, swinging the verdict towards the other class, C1. The only features increasing the likelihood or probability of C2 being the correct label are F9, F8, and F4. Given that only few features positively contribute to arriving at the C2 prediction, it is very strange that the model has 100.0% confidence in its prediction for the selected instance.

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

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

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

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

High

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

RandomForestClassifier

C2

Music Concert Attendance

There is an 80.0% chance that the true label for the given case is C2. Nine out of twenty features have a positive impact. Most features have a moderately low positive or negative impact, with the exception of F2, F11, and F5 and it appears as if F2 has an extremely negative impact, while F11 and F5 have the greater positive impacts. F15 has positive impacts, whereas the attributions of the features F13 and F18 are negatives. The least important features include F16, F8, F9, F14, F4, F7, and F19 with varying smaller effects.

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

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

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

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

< 10k

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

RandomForestClassifier

C1

Employee Attrition

There is disagreement about which label is acceptable for the case under consideration since the model is unsure which of the two labels is right. The confusion in the aforementioned classification may be attributable only to the effect of F8. F8 is by far the most influential variable, with a negative contribution that reduces the chance of label C1 being the correct label in the given case substantially; supporting the that case should be labelled as C2. Compared to the influence of F8, the remaining variables have a moderate to low effect on the classification decision made here for the case under consideration. F16, F9, and F1 are notable moderately key variables, with positive contributions boosting the likelihood of label C1. F7, F21, F26, F3, F23, F15, F13, F20, F22, and F10 are not among the features demonstrated to contribute to the classification above; since they have very insignificant impact on the model's conclusion here.

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

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

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

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

Stay

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