Researchers at the University of Chicago have created an algorithm that forecasts crime weeks in advance. This occurs within a radius of 300 meters with 90 percent accuracy.
The work was released in a journal called Nature. The system assimilates patterns from publicly available data on crimes that involve violence and property. According to the researchers such events are less susceptible to police prejudice. The model separates crime by looking at the temporal and spatial coordinates of individual events and identifying patterns to forecast future occurrences.
The algorithm splits the city into “spatial tiles” about 300 metres wide and forecasts violations in these areas. According to the researchers, previous models depended more on traditional neighborhood or political boundaries, which are liable to bias. Ishanu Chattopadhyaya, the lead author of the study, stated that the tool’s high accuracy does not mean that it should be used in law enforcement policy management. Additionally, He said that police departments shouldn’t make use of it for preventive intrusion into neighborhoods to stop crime. He said, “This is not magic; there are limitations but we tested it, and it works very well.”
Instead, the algorithm should be attached to the toolkit of urban policy and crime-fighting police strategies. According to the scientist “We have created a digital twin of the urban environment. If you give it data about what happened in the past, it will tell you about the future.” The team also looked into the police response to crime by examining the number of arrests after incidents and comparing rates across neighborhoods.