MarceloLZR/titanic

Titanic (Survived/Not Survived) - Binary Classification

How to use

from huggingface_hub import hf_hub_url, cached_download
import joblib
import pandas as pd
import numpy as np
from tensorflow.keras.models import load_model

REPO_ID = 'danupurnomo/dummy-titanic'
PIPELINE_FILENAME = 'final_pipeline.pkl'
TF_FILENAME = 'titanic_model.h5'

model_pipeline = joblib.load(cached_download(
    hf_hub_url(REPO_ID, PIPELINE_FILENAME)
))

model_seq = load_model(cached_download(
    hf_hub_url(REPO_ID, TF_FILENAME)
))

Example A New Data

new_data = {
    'PassengerId': 1191,
    'Pclass': 1, 
    'Name': 'Sherlock Holmes', 
    'Sex': 'male', 
    'Age': 30, 
    'SibSp': 0,
    'Parch': 0, 
    'Ticket': 'C.A.29395', 
    'Fare': 12, 
    'Cabin': 'F44', 
    'Embarked': 'S'
}
new_data = pd.DataFrame([new_data])

Transform Inference-Set

new_data_transform = model_pipeline.transform(new_data)

Predict using Neural Networks

y_pred_inf_single = model_seq.predict(new_data_transform)
y_pred_inf_single = np.where(y_pred_inf_single >= 0.5, 1, 0)
print('Result : ', y_pred_inf_single)
# [[0]]