bgarzonm/ml-uifce

Customer Churn Prediction Project

Overview

This GitHub repository contains the code and resources for Customer Churn Prediction project. The goal of this project was to develop a predictive model that can identify potential customer churn.

Project Highlights

• Explored and analyzed the dataset.
• Preprocessed the data, encoding categorical features, and scaling numerical features.
• Utilized various machine learning algorithms for classification, including:
  • XGBoost
  • Random Forest
  • Logistic Regression
• Employed deep learning.

Repository Structure

• data: This directory contains the dataset used for the project.
• features: Jupyter notebooks used for EDA.
• models: Jupyter notebooks used for model training, and evaluation.

Usage

1. Clone this repository to your local machine.

   git clone https://github.com/bgarzonm/ML-UIFCE.git
   

2. Navigate to the models or features directory to explore the Jupyter notebooks

3. Feel free to adapt the code and models for your own use case.

Results

After training and evaluating the models, we achieved the following performance:

• XGBoost:

  • Accuracy: 0.99
  • F1-Score: 0.98

• Random Forest:

  • Accuracy: 0.98
  • F1-Score: 0.97

• Logistic Regression:

  • Accuracy: 0.89
  • F1-Score: 0.88

Conclusion

The results from our models can help businesses identify customers at risk of churning and take proactive measures to retain them, ultimately improving customer retention and business profitability.