Data Fusion Algorithms for Mapping Crime

Topanga, also known as Division 21, is a new division of the Los Angeles Police Department (LAPD).  One of the major problems the Topanga Division faces is burglary from motor vehicle (BFMV), and one of the ways the LAPD seeks to address the problem is through crime mapping.   Crime mapping is part of a broader initiative to utilize data to inform and direct police work, commonly known as Smart Policing Initiative (SPI). This project is a particular application of the SPI philosophy to manage BFMV crime in the LAPD Division of Topanga.   The goal of the project is to improve hot-spot mapping techniques through data fusion algorithms to produce relative probability density estimates of crimes occurring in a given region.  We develop and improve upon algorithms that make use of Maximum Penalized Likelihood Estimation and Kernel Density Estimation methods to produce hot-spot maps.  We compare existing algorithms against our modified versions to determine which have better predictive capabilities.  Upon further analysis and experimentation we expect our algorithms will be able to create improved predictive hot-spot maps.  An increased understanding of the distribution of these crimes in the Topanga area may serve the LAPD in reducing the occurrence of BFMVs and focusing valuable resources on problem areas.