Criminal behavioral profiling tool based on FBI open data.
Law enforcement and criminologists need tools to identify behavioral patterns from crime data. Virtual Profiler leverages machine learning to build predictive profiles from real FBI datasets.
- Python — core language
- Scikit-learn — predictive modeling
- Pandas / NumPy — data processing
- Streamlit — interactive web app
- FBI Crime Data API — open data source
- Behavioral pattern clustering
- Crime type prediction by geographic zone
- Interactive profiling dashboard
- Explainability layer (SHAP values)
🟡 In progress — active development
Chams-Eddine Mokdad — Lead Data Scientist
LinkedIn ·
Portfolio