Customer Churn Prediction for a Telecom Company

Objective: Build a machine learning model to predict which customers are likely to leave the company, and identify key factors driving churn.

? Tools & Tech:

Python (Scikit-learn, Pandas, NumPy)

Jupyter Notebook

Tableau or Power BI for visualization

Dataset Includes:

Customer demographics

Service usage (minutes, data, SMS)

Subscription type

Tenure

Churn label (Yes/No)

Workflow:

Data cleaning and feature engineering

Exploratory data analysis to find churn patterns

Train/test split and model selection (e.g., logistic regression, random forest)

Evaluate model performance (accuracy, precision, recall)

Visualize churn risk by customer segment

Outcome: Model achieved 85% accuracy. Found that customers with short tenure and high complaint frequency were most likely to churn. Recommended targeted retention campaigns.