Department of Biostatistics & Medical Informatics Seminar with Muxuan Lian of MD Anderson Cancer Center
Gradient-Augmented AI Calibration for Covariate Adjustment in Randomized Trials
Event Details
Date
Friday, October 16, 2026
Time
Noon-1 p.m.
Location
Description
Dr. Liang proposes GRACIAS, which combines repeated AI-generated outcome predictions based on randomly sampled examples/demonstrations with low-rank features obtained from model gradients. The simulations and analyses of 21 randomized multi-arm trials connect flexible AI prediction with classical regression adjustment and provide a principled framework for using AI-derived information to improve treatment-effect estimation.
Cost
Free
Contact
Accessibility
We value inclusion and access for all participants and are pleased to provide reasonable accommodations for this event. Please email juan.caicedo@wisc.edu to make a disability-related accommodation request. Reasonable effort will be made to support your request.