Data and machine learning case study
Predictive Crime Analysis
Datathon project | 2024
This project combined exploratory data analysis, visualization, and machine learning to understand crime behavior at scale. The goal was to turn raw public data into clearer operational signals for deployment planning.
Outcome
Produced 15+ actionable insights and a predictive model for hourly crime risk forecasting.
Selected implementation work
What I built and improved
- Processed 500k+ records with Python and Power BI to uncover spatial and socio-economic crime patterns.
- Delivered 15+ insights spanning crime types, demographics, and temporal trends.
- Built a machine learning model to predict hourly crime rates and highlight high-risk periods.
Technology
Tools and capabilities
- python
- power-bi
- machine learning
- data analysis
- tensorflow
- scikit-learn
Explore