Machine Learning

Machine learning tools continually leverage data to “learn” and improve performance — whether that’s cleaning datasets or analyzing the data within them to make recommendations.

Good data is essential to developing successful interventions to reduce violence and reform our criminal justice system. But too often, public safety datasets are disjointed or otherwise incomplete, which makes it difficult to analyze the effects of an intervention. Machine learning tools can strengthen data analysis by gathering information across multiple datasets or by making predictions based on trends in the data — allowing researchers to analyze data more efficiently.  

The Crime Lab team is utilizing and developing machine learning tools to expand and improve our data analysis capacity within many of our projects. 

Latest Updates

Second Chance Month: Centering Lived Experience in Violence Intervention
Media Mention
National League of Cities
Apr 2024

Second Chance Month: Centering Lived Experience in Violence Intervention

In recognition of Second Chance Month, the National League of Cities’ Maryam Ahmed and Kirby Gaherty write about the importance of centering “credible messengers”—people with lived experience in the justice system—to advance community safety and highlight the Crime Lab’s Community Violence Intervention Leadership Academy.

Why America fell for guns
Essay
Aeon
Apr 2024

Why America fell for guns

Megan Kang, a Crime Lab affiliate and Ph.D. candidate in sociology at Princeton University, writes an essay that describes America’s extraordinary levels of gun ownership in the context of a key turning point in US gun culture in the mid-20th century.

Oeindrila Dube on Cognitive Behavioral Training for Police
Podcast
Probable Causation
Apr 2024

Oeindrila Dube on Cognitive Behavioral Training for Police

In this episode of Probable Causation, Dr. Oeindrila Dube discusses her research on Situational Decision-Making, a cognitive behavioral training program for police.