Thomas is a member of the FHI DPhil Scholars program, and is currently completing a DPhil in Computer Science. He is particularly interested in bridging the gap between practical, less-understood machine learning algorithms, and well understood but sometimes overly simplistic or impractical machine learning theory. Current efforts towards this direction include understanding how to do inference in time constrained environments, and explicitly modeling concepts like “teaching learning” and “meta learning” in a machine learning theory context. His thesis title is currently “Machine Learning Tools for Lifelong Learning”.
Thomas completed an undergraduate degree in Computer Science and Mathematics, and a masters degree in Statistics, at Harvard University.