Mathematics of Crime (Andrea Bertozzi - UCLA)

4:00 pm on Thursday, February 28, 2019
5:00 pm on Thursday, February 28, 2019
MCS 148, 111 Cummington Mall
I will review research at UCLA related to the modeling and analysis of crime patterns, most notably crime hotspots and self-excitation of crime,, in particular gang crimes. The models include residential burglaries using PDEs, stochastic self-exciting point process models for repeat victimization and gang retaliation, and deep learning for real time prediction. The point process work has led to a predictive policing software program that is in use in over 50 cities worldwide. We have also been analyzing videos from police body worn cameras using machine learning methods for semi-supervised learning.