End-To-End SK-Learn pipeline for predicting survival on the titanic using logistic regression model.
In this pipeline, we are using custom transformers to preprocess the data and using LogisticRegression estimator to train and predict the results.
1- Feature Imputation:
- MeanMedianImputer
- CategoricalImputer
- AddMissingIndicator
2- Feature Encoding:
- OneHotEncoder
- RareLabelEncoder