GitHub Star Predictor AI 🚀

A machine learning pipeline that fetches live repository data from the GitHub API, engineers dataset features, and trains a Linear Regression model to predict repository popularity (stars).

Project Structure

• fetch_data.py: Connects to GitHub API and downloads the top 100 repositories.
• clean_data.py: Cleans raw strings and engineers a feature calculating repository age in days.
• train_model.py: Splits data, trains a Scikit-Learn Linear Regression model, and tests its accuracy.
• app.py: Launches an interactive Streamlit UI web app to test custom inputs against the AI model live.

Performance

• Model Type: Linear Regression
• R² Accuracy: 0.58 (Explains 58% of the variance)
• Sample Prediction: A repo with 5,000 forks is predicted to reach ~61,000 stars!