The AI ​​model that predicts crime in US cities is right nine times out of 10

An AI model correctly predicted crimes a week before they happened with 90% accuracy in eight US cities, the model’s co-creator said.

Ishanu Chattopadhyay, an assistant professor at the University of Chicago, told Insider that he and his team created an “urban twin” to monitor crime data in Chicago from 2014 to late 2016 before predicting the likelihood of certain crimes for the following weeks. , with 90% accuracy within a two-block radius.

The model, which had similar results in seven other cities, focused on the types of crimes committed and where they occurred. Chicago’s crime rate in 2020 was 67% higher than the national average, according to data compiled by AreaVibes.

According to research compiled by Econofact, racial bias in policing has high economic costs and worsens inequality in areas already experiencing high levels of deprivation.

While some models attempt to eliminate these biases, they have often had the opposite effect, with accusations that racial biases in the underlying data reinforce future biased behaviors.

In 2016, the Chicago Police Department tested a model to predict those most at risk of being involved in a shooting, but the secret list ultimately revealed that 56% of black men living in the city were on the list. , fueling accusations of racism.

Chattopadhyay said their model found that arrests increased alongside reported crimes in high-income neighborhoods, while arrests were stable in low-income areas, suggesting some bias in police response.

“We demonstrate the importance of uncovering city-specific patterns for predicting reported crime, which generates a new view of city neighborhoods, allows us to ask new questions, and allows us to evaluate the action of policing in new ways,” co-author James Evans told Science Daily.

Lawrence Sherman of the Cambridge Center for Evidence-Based Policing told New Scientist he was concerned about the inclusion of police data in the study that depended on reports from citizens or the crimes the police were looking for.

Chattopadhyay agreed that this was a problem and that his team had tried to account for it by excluding crimes reported by citizens and police interventions, usually involving minor drug offenses and traffic stops. , and zoning out more serious violent and property crimes that were more likely to be reported in any setting.

Chattopadhyay, who made the data and algorithm publicly available to increase scrutiny, hoped the results would be used for high-level policy and not as a reactive tool for police.

“Ideally, if you can predict or anticipate crime, the only answer is not to send in more officers or flood a particular community with law enforcement,” Chattopadhyay said. “If you could prevent crime, there are a host of other things we could do to prevent such things from actually happening so no one goes to jail and help communities as a whole.”

About Geraldine Higgins

Check Also

Chargebee empowers subscription companies to fight

San Francisco, Calif., July 28. 2022 (GLOBE NEWSWIRE) — Chargebee, the leading subscription management platform, …