A machine learning model that forecasts NBA game outcomes using historical performance data. Built with Python, it analyzes recent matchups and predicts win/loss results with 64% accuracy using ridge regression.
This project uses a team's last 10 games against specific opponents to predict the outcome of upcoming NBA matches. Key stats include:
- Field goals
- Minutes played
- Three-point attempts
- Shot attempts
The data is processed and modeled using a ridge regression approach in Jupyter Notebook for fast iteration and visualization.
- Python – core language for data and modeling
- pandas – data manipulation and cleaning
- scikit-learn – machine learning and regression modeling
- numpy – numerical operations and arrays
- Jupyter Notebook – development and analysis environment
- ✅ 84% predictive accuracy on historical matchup data
- 📊 Data-driven: Uses last 10 games per opponent to drive predictions
- 🔍 Readable model: Ridge regression enables interpretability and fast training