The Limits of Transfer Learning

The Walter Bradley Center Computer Science, 2019-20

Liaison(s): Robert J. Marks
Advisor(s): George Montañez
Students(s): Jake Williams (PM-S), Tyler Sam (PM-F), Abel Tadesse, Huey Sun

The Walter Bradley Center Clinic team is conducting research on a branch of machine learning called transfer learning, where insight from one problem is transferred to solve or get a head start on another related, more difficult problem. Our goals are to develop bounds on the efficacy of transfer, and our deliverables include academic papers, theorems and explanations about the nature of our work.