This project aims to predict the likelihood of a customer churning using a logistic regression model. The model is trained using customer data, including demographics and usage statistics, and can make predictions for individual customers or batches of customers. The model is deployed using Streamlit, and can be accessed through the following link: https://prathamsoneja-willtheystay-home-lrv0gl.streamlit.app/
The data used in this project is available on Kaggle and can be found here. The data includes information on customer demographics and usage statistics, such as account length, number of voice mail messages, and total night charge.
A logistic regression model is trained using the data to predict the likelihood of a customer churning. The model is trained in a Jupyter notebook, which can be found at this link
The model is evaluated on a holdout set and achieves an accuracy of 81.17%.
The Streamlit app allows users to input information for a single customer and receive a prediction of the likelihood of that customer churning. The app also allows users to upload an Excel file with information on multiple customers, and receive predictions for each customer in the file.
Pull requests and suggestions for improvement are welcome.
This project is licensed under the Mozilla Public License - see the LICENSE.md file for details
Kaggle
Streamlit
Jupyter Notebook