Customer Churn Prediction

This project aims to predict customer churn using the Kaggle Telco Customer Churn dataset. Various machine learning models, including decision trees, random forests, support vector machines, and neural networks, are employed to analyze customer churn in the telecommunications sector.

Project Summary

In this study, various machine learning and data analytics methods were used to improve customer churn prediction. The Kaggle Telco Customer Churn dataset provides comprehensive information about customer behaviors. The primary goal is to develop a model that can predict potential customer churn in advance, enabling businesses to offer targeted customer retention programs.

Dataset and Features

The dataset includes the following features:

• Churn: A binary variable indicating customer churn.
• Services: Information about phone services, internet connection, security features, etc.
• Account Information: Contract types, payment methods, and billing preferences.
• Demographic Information: Gender, age range, and family structure.

Models Used

• Decision Tree
• Random Forest
• Support Vector Machine
• Neural Network

Project Files

• exploratory_data_analysis.py: Exploratory data analysis.
• neural_network.py: Training and evaluating the neural network model.
• decision_tree.py: Creating and evaluating the decision tree model.
• random_forest.py: Creating and evaluating the random forest model.
• svm.py: Creating and evaluating the support vector machine model.