Custmer Churn Prediction

About the Project

This is a Customer Churn Prediction application that predicts whether a customer is likely to churn based on historical data. The application follows a robust pipeline-based workflow for preprocessing, model building, and evaluation. It also includes a Flask web application for easy user interaction and is containerized using Docker for seamless deployment and scalability.

Workflow

The application workflow is structured as follows:

1. Data Ingestion:

   • Load the dataset from a CSV file.

2. Data Cleaning and Preprocessing:

   • Handle missing values.
   • Encode categorical variables.
   • Normalize/standardize numerical features.

3. Pipeline Creation:

   • Build a pipeline using scikit-learn to streamline preprocessing and model training.
   • Save the pipeline using joblib for reuse.

4. Model Building and Evaluation:

   • Train multiple classification models.
   • Evaluate models using metrics such as accuracy, precision, recall, and F1-score.
   • Select the best-performing model.

---

Key Features

• End-to-end machine learning pipeline for classification tasks.
• Modular and reusable code for data preprocessing and model training.
• Simple and intuitive web interface built with Flask.
• Fully containerized setup for deployment using Docker.