Customer Churn Prediction

Mission of the Project

To predict customer churn using machine learning so that businesses can improve retention and reduce losses. By identifying at-risk customers early, companies can take proactive steps to improve satisfaction and reduce revenue loss.

Introduction

Customer churn refers to the percentage of customers who stop using a company's services. This project uses machine learning to analyze patterns and predict which customers are likely to churn, enabling businesses to intervene before it happens.

Problem Statement

Many companies struggle to identify customers who are likely to leave. Without prediction, businesses lose revenue and spend more on new customer acquisition — which costs 5–7× more than retaining existing customers. Timely identification is crucial.

Key Features

Churn prediction with ML algorithms

Data preprocessing & feature engineering

Exploratory data analysis (EDA)

Accuracy, Recall & F1-Score evaluation

Visualization of churn insights

Logistic Regression & Random Forest

How It Works

Dataset Loaded Data Cleaned Feature Selection Train-Test Split ML Model Trained Metrics Evaluated

Tools & Technologies

Python

Pandas

NumPy

Scikit-learn

Matplotlib & Seaborn

Project Repository

View on GitHub