Exercise: Handling imbalanced data in machine learning

1. Use this notebook but handle imbalanced data using simple logistic regression from skelarn library. The original notebook using neural network but you need to use sklearn logistic regression or any other classification model and improve the f1-score of minority class using,

   1. Undersampling

   2. Oversampling: duplicate copy

   3. OVersampling: SMOT

   4. Ensemble

   Solution

2. Take this dataset for bank customer churn prediction : https://www.kaggle.com/barelydedicated/bank-customer-churn-modeling

   1. Build a deep learning model to predict churn rate at bank

   2. Once model is built, print classification report and analyze precision, recall and f1-score

   3. Improve f1 score in minority class using various techniques such as undersampling, oversampling, ensemble etc

   Solution

   Thanks https://github.com/src-sohail for providing this solution.