AI-SDM Student Brainstorming Session
October 1, 2026 12:00PM—1:00PM
Location:
In Person and Virtual - ET
-
Wean Hall 1327 and Zoom
Topic
Peer review at major research journals now involves hundreds of submissions per month, and the case for automated reviewer assignment is clear. What is less clear is how to do it well: unlike conference assignment, which can be solved offline once all submissions are in, journal assignment must be done online: each paper assigned immediately upon arrival, from a pool of reviewers who may become unavailable for weeks once assigned.
We study this as an instance of online task assignment with reusable resources: tasks arrive over time, each must be immediately assigned to k of R available resources, and each assigned resource is busy for d subsequent timesteps. This model captures peer review, healthcare scheduling, ridesharing, and emergency dispatch. I will present restless-bandit based policies and upper bounds. I will also discuss a closed-form Whittle index policy for automated reviewer assignment and insights from evaluation on a four-year trace of submissions to TMLR.
Led by students of all levels, these sessions offer a casual forum to discuss AI-related research while exploring cross-cutting connections between various disciplines in artificial intelligence. All students welcome!
REGISTER ⇢ Register to attend in-person or on Zoom
For More Information:
pwerns@andrew.cmu.edu