Customer churn is a major worry for large corporations, particularly in the telecommunications industry, because it directly affects revenues. Having a way in which these companies can better understand why their customer churn allows them to save themselves regarding revenues and even reputation. This web application uses the churn prediction model that assists telecom operators in predicting customers who are most likely subject to churn.
The application for this web application: https://customerchurnmlp.streamlit.app/
Video Demonstration link: https://drive.google.com/file/d/1EKOW08gX_zqreDJDKBY9hQ85eUj6B0Ur/view?usp=sharing
The online application collects information using eleven features that allow users to enter their choices or select options from dropdown menus directly. Users can begin the prediction process by clicking the "Predict" button after selecting their selections. It is critical to understand that, for senior citizens in particular, a number of 1 correlate to "Yes," while a value of 0 corresponds to "No."
The building and training of the model made use of the following:
• The CustomerChurn_dataset • Streamlit: A library for creating interactive web applications with Python. • NumPy: A library for numerical operations in Python. • Pandas: A data manipulation and analysis library for Python. • TensorFlow: An open-source machine learning framework for building and training neural networks. • Scikit-learn: A machine learning library for simple and efficient data mining and analysis tools.
Please Note: This does not mean these are the only libraries used in the project.