The strategy includes taking crime data and advanced analysis to determine where and when crime could occur. With that forecast, police can devise strategies and assign officers to prevent crime, officials said.
The strategy allows police to “take limited resources and employ them in such a way to improve the overall safety of the community,” Glendale Police Chief Ron De Pompa said.
“It's kind of new for us,” he said. “We are experimenting with this implementation currently and realizing some significant success.”
Glendale police applied the strategy in May to certain neighborhoods in South Glendale and saw a 78% reduction in property and violent crimes the following month, Povilaitis said.
Predictive policing allows officers to identify and track repeat criminals, and examine areas with high crime activity, he added.
Police analyze five years of data for a certain neighborhood to “determine historic peak hourly time periods for crime,” according to a city report.
UCLA and Santa Clara University, along with the Santa Cruz Police Department, has been examining the policing strategy. Researchers applied algorithms that were used for predicting aftershocks to Santa Cruz's crime data and assigned appropriate police resources, Povilaitis said.
Various strategies, including predictive policing, have contributed to Glendale's comparatively low crime ranking, De Pompa said.
“That doesn't happen by accident,” he said.
Glendale was ranked 10th last year among cities with the lowest crimes rates for populations of 100,000 to 499,000, according to CQ Quarterly rankings released in December. That report used data from the FBI's 2010 crime statistics.