Analyzing Crime in Philadelphia using Bayesian Clustering

Course Project
Spatial Statistics
Course project looking at Bayesian clustering applied to studying crime in Philadelpha
Published

May 2, 2026

Source Code Read Paper Slides

This project was completed in Spring 2026 as part of a PhD-level advanced spatial statistics course. This project contained two parts:

  1. A group presentation reviewing the paper “Methods and Simulation Study Case Study and Results References Crime in Philadelphia: Bayesian Clustering with Particle Optimization” studying the methods utilized and reproducing the results. The presentation is attached above.
  2. An individual data analysis either expanding or improving upon the results. My approach here was to simplify the methodology used and do a fully Bayesian cluster analysis using MCMC. This produced interpretable results with efficient computation and full uncertainty quantification. The brief report with the results is attached along with the file with the code in the GitHub repository.