San Francisco Crime Data Mining & Prediction

-Team Project – Data Mining & ML | Python, Scikit-learn, Streamlit-

•Clustered and analyzed 878,000+ crime records, identifying 5+ major spatial crime zones to support predictive modeling with 85%+ model accuracy.

•Developed Decision Tree & Random Forest models for crime category prediction •Performed feature engineering, outlier removal, and feature encoding

•Evaluated models using Silhouette Score, Accuracy, and Confusion Matrix