Churn (loss of customers to competition) is a problem for companies because it is more expensive to acquire a new customer than to keep your existing one from leaving. This problem statement is targeted at enabling churn reduction using analytics concepts. The objective of this case is to predict customer behaviour.
It is a Classification Problem.
All the steps implemented in this project
Data Pre-processing.
Data Visualization.
Outlier Analysis.
Missing value Analysis.
Feature Selection.
Correlation analysis.
Chi-Square test.
Feature Scaling.
Standaridization.
Dealing with Target Class Imbalance in the Dataset & Splitting into Train and Test Dataset.
What is imbalance in data.
What is target class imbalance problem
Sampling techniques to overcome imbalance:
Random Under Sampling.
Random Over Sampling.
Using Both Under and Over Random sampling.
Synthetic Minority Over Sampling Tehnique.
Model Development
Decision Tree.
Random Forest.
Logistic Regression.
KNN.
Naive Bayes.
Model Performance.
Conclusion
R code.
Python Code.
GitHub folder contains:
R code of project in ‘.R format’: Churn Reduction using R.R
Python code of project in ‘.ipynb format’: Churn Reduction using Python.ipynb