Customer Churn Prediction Using ANN

This project focuses on predicting customer churn using an Artificial Neural Network (ANN). The model was developed using TensorFlow/Keras and deployed as a user-friendly web application via Streamlit. The solution leverages robust preprocessing pipelines to handle data efficiently, enabling businesses to identify and retain at-risk customers.

Features

• Customer Churn Prediction: Predict whether a customer is likely to churn based on input features.
• Interactive Web App: User-friendly interface to input data and get instant predictions.
• End-to-End Preprocessing: Automated handling of missing values, normalization, and encoding of categorical variables.
• Model Deployment: Hosted on Streamlit for easy accessibility.

Hosted Link

Access the application here: ANN Churn Prediction Web App

Technologies Used

• TensorFlow/Keras: For building and training the ANN model.
• Streamlit: For creating an interactive and accessible web application.
• Pandas & NumPy: For data preprocessing and feature engineering.
• Scikit-learn: For splitting data and preprocessing utilities.

Project Overview

1. Data Preprocessing:
   • Handled missing values and performed feature scaling.
   • Encoded categorical variables using one-hot encoding.
2. Model Development:
   • Designed a feedforward neural network with multiple dense layers.
   • Optimized the model using backpropagation and various hyperparameters.
   • Evaluated performance using metrics like accuracy and recall.
3. Deployment:
   • Deployed the trained model as a Streamlit app for real-time predictions.

Installation

To run this project locally:

1. Clone the repository:

   git clone https://github.com/krishnaik06/ANN-CLassification-Churn.git
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Run the Streamlit app:

   streamlit run app.py
   

Usage

• Open the hosted app or run it locally.
• Enter customer details in the input fields.
• Click on the "Predict" button to get the churn prediction.

Future Enhancements

• Incorporate additional features to improve prediction accuracy.
• Add a feature to upload bulk data for batch predictions.
• Enhance the user interface for better user experience.

License

• This project is licensed under the MIT License.