During the fall semester, in-person concerts, events and lectures that involve outside guests will not be held, per the university’s COVID-19 travel and visitor policy.
This is a past event. Its details are archived for historical purposes.
The contact information may no longer be valid.
Please visit our current events listings to look for similar events by title, location, or venue.
Wednesday, September 16, 2020 at 1:15pm to 2:30pmVirtual Event
"The Lessons and Limits of Predicting Shooting Victimization" (with Sara Heller, Benjamin Jakubowski, and Zubin Jelveh)
Abstract: Risk prediction algorithms have the potential to reduce gun violence by improving how prevention efforts are allocated. But using algorithms responsibly requires understanding what drives predictions and errors, whom they identify and miss, and why. This paper uses Chicago Police Department data to predict a person's risk of being shot using a machine learning algorithm, reports on its sensitivity to important modeling choices, and quantifies out-of-sample performance both overall and across demographic groups. The model identifies a small group of people with a risk of becoming shooting victims far greater than that faced by the average Chicagoan. But it also highlights the need for other data sources and broader prevention efforts: The model fails to identify the vast majority of eventual victims, missing female victims almost entirely. Differing causes of violence and differing representation in police data likely help explain differences in predictive power across race and age groups.
PAM virtual seminar. Please email the event coordinator for the zoom link.