Customer retention is critical for service providers. Predicting churn allows companies to take proactive steps to retain customers, which can lead to significant savings and increased profitability. Churn prediction helps in identifying customers at risk, enabling service providers to improve customer satisfaction and loyalty.
This project involves a comprehensive data analysis process divided into five main sections. Each section is managed by a specific collaborator responsible for its successful execution. Below is a detailed explanation of each section along with the respective collaborator.
This section involves gathering data from reliable sources and exploring it to get an initial understanding. For this project, we took the Telco Customer Churn dataset from Kaggle. The steps include calculating basic summarization statistics, visualizing the data, and identifying data problems such as outliers, missing data, and duplicate records.
- Responsible Collaborator: F. Mulhima
- Contact: [email protected]
This section focuses on cleaning the data by removing inconsistencies or errors and handling missing values. The steps include binning, smoothing, regression, clustering, and dimensionality reduction.
- Responsible Collaborator: N. Abisha
- Contact: [email protected]
This section continues data preprocessing by selecting attribute subsets, performing numerosity reduction (sampling or modeling), and transforming the data into a suitable format for analysis. The steps include normalization, feature selection, feature engineering, discretization, and concept hierarchy generation.
- Responsible Collaborator: T. F. Fathima
- Contact: [email protected]
This section involves applying data mining methods/tasks and evaluating the resulting models/patterns. It includes experimenting with different parameter settings and multiple alternative methods, improving preprocessing and feature generation, and increasing the amount or quality of training data.
- Responsible Collaborator: H. M. F. Asma
- Contact: [email protected]
This section presents the analysis findings and validates the models' performance and reliability. It covers performance metrics, cross-validation, model selection, hyperparameter tuning, and addressing overfitting and underfitting issues.
- Responsible Collaborator: G. Kishon
- Contact: [email protected]
We have deployed our project as ChurnX on Streamlit Cloud, where you can interact with the application and explore its features. Access it here.
Contributions are welcome! Feel free to fork this repository, make improvements, and submit pull requests. For major changes, please open an issue first to discuss what you would like to change.
This project is licensed under the MIT License - see the LICENSE file for details.