ML_GDG

• Code for the GDG Machine Learning Talk on "Machine Learning for Imbalanced Class Distributions".
• Contains implementation of data sampling and algorithms that were discussed in the session for the Customer Churn Prediction.

Data Sampling Methods:

1. Over Sampling
2. Under Sampling
3. SMOTE

Algorithms:

1. Random Forest Classifier (with and without class weights)
2. Support Vector Machines (with and without class weights)
3. Neural Networks (Fully Connected)
4. Cost Sensitive Learning (Logistic Regression and Random Forest Classifier)