Mathematical Model Predicts Burglary Hotspots

Math isn't just limited to the books, in fact, it can help fight crime in the streets. A team of researcers have developed a mathematical model that predicts burglary patterns.

Criminal activity has been recently noticed as having a clustered, or "hot-spot," characteristic to it. Crimes tend to aggregate in space and time in urban settings. In an effort to better understand these criminal patterns, experts are increasingly turning to mathematical models for an explanation.

Houses that have burglarized before or are close neighbors of houses that have been burglarized are at a higher risk of burglary. To help predict the pattern in which these hotspots occur, the authors  Steve Cantrell, Chris Cosner, and Raúl Manásevich propose a mathematical model.

"Our research provides a mathematically rigorous way of connecting the geographical characteristics of a neighborhood [such as demographics, economics and ecology] to the patterns of burglary that would be seen in the neighborhood," says lead author Steve Cantrell. 

"Bringing geography into the model is an important step in understanding the model in realistic situations."Our work was inspired by models of burglary patterns that were developed by a group of mathematicians and scientists at UCLA," he goes on to explain. 

The UCLA group analyzed the dynamics of burglary hotspots based on the assumption that criminal agents strike based on a house's "attractiveness value."

The "attractiveness value" is ratio of payoff to the negative consequences or chances of being caught. If a house has been burglarized before, it will have a higher "attractiveness value" since someone has already gotten away with it in the past.This creates a "broken window effect," or a repeat victimization sociological effect.

By using one model based on burglar movement patterns, and the other based on the attractiveness of a house, the researchers were able to come up with a model that predicts burglary hotspots.