Dr. Jelveh joined the College of Information Studies at the University of Maryland (UMD) in January 2021 with a joint appointment in the Department of Criminology and Criminal Justice. Prior to arriving at UMD, Dr. Jelveh served as research director in Crime Lab’s New York office, working with government and community partners in New York, New Jersey, and Chicago.
Dr. Jelveh works on developing and evaluating prediction models in the domains of domestic violence, gun violence, behavioral health, and pretrial justice. In support of that work, Dr. Jelveh also develops record linkage algorithms tailored for administrative data. Finally, Dr. Jelveh studies the science of science, in particular the incentives that generate p-hacking and the effectiveness of approaches for curtailing it.
Machine Learning Can Predict Shooting Victimization Well Enough To Help Prevent It
This National Bureau of Economic Research Working Paper shows that shootings are predictable enough to be preventable.
New Jersey Portfolio
The Crime Lab is partnering with the New Jersey Administrative Office of the Courts (AOC) to help strengthen reforms to the state’s criminal justice system introduced in 2017, including eliminating cash bail and introducing a risk assessment tool to aid in pretrial release decisions.
Targeted Interventions for High-Risk Domestic Violence Victims
The Crime Lab studied targeted approaches to identify high-risk domestic violence victims, conduct outreach, and provide resources to lower the risk of experiencing future harm.
Service Provision Risk Assessment (SPRA)
The Crime Lab designed the Service Provision Risk Assessment to help violence prevention organizations identify the people at the highest risk of future gun violence involvement and offer them intensive social services and supports.
New York City Domestic Violence Analysis
The Crime Lab partnered with the New York City Police Department (NYPD) to identify New Yorkers at high risk of domestic violence to ensure those individuals received social services.
