Researchers develop predictive policing model for Santa Cruz PD

The Santa Cruz Police Department has adopted a new program in predictive policing that uses mathematical modeling developed by a research team led by the UCLA Department of Mathematics. The model predicts “hotspots” where future crimes, such as burglaries and car thefts, are likely to occur in the same way that earthquake models predict aftershocks.

 

In Santa Cruz, law enforcement deployed patrols to targeted crime hotspots identified by the mathematical model, which resulted in a 27 percent drop in burglaries in July compared to a year ago. The model is based on the last eight years of Santa Cruz crime history and is updated daily to create a list of 10 hotspots where crimes are most likely to occur, directing police patrolling efforts to prevent crimes from escalating. Other cities, including Boston and Chicago, are considering adopting predictive policing as law enforcement agencies contend with scarce resources and budget cuts.

 

The academic research team that created the model included UCLA faculty members Martin Short, assistant adjunct professor of mathematics, Jeffrey Brantingham, associate professor of anthropology, and Frederic Paik Schoenberg, associate professor of statistics, as well as researchers from Santa Clara University and UC Irvine.