Project Description

This project is based on predicting churning behaviour of customer of a telecom company.
we need to predict either customer will churn out or not. This is classification problem.

Prerequisites

In Python, Make sure in your enviornment below packages are installed.

numpy, pandas, matplotlib, seaborn, sklearn, imblearn

In R, make sure below packages are installed:

ggplot2, corrgram, corrplot, randomForest, caret, class, e1071, mlr, grid, DMwR, irace, usdm

Files Detail

• Train_data.csv and Test_data.csv containing dataset first will be used in builiding model second will be used in predicting and testing our model
• churn_prediction.py contain implemenation in Python
• churn_prediction.R contain implementation in R
• project_report_explained.pdf contain Process and anlysis done during this project

Highlights

• SMOTE upsampling Used for Target imbalance case
• K-fold CV used for comaparing models
• Hyperparameter tuning used for improving preformance of models

Note: make sure to change path as per file location