I am a Ph.D. student at Simon Fraser University advised by Manolis Savva and Sharan Vaswani. Prior to this, I worked as a full-time Data Scientist in the San Francisco Bay Area for 3 years. Before that, I was a research and teaching assistant at the University of British Columbia working in Applied Mathematics. I received my Bachelor’s degree in Pure Mathematics from the University of Toronto.
- My current research is in the field of Optimization in Machine Learning. More specifically, I’m interested in optimization algorithms that only use first-order derivatives but can interpolate between first and second-order methods. Two recent works in this area are LocoProp and Target-based Surrogates for Stochastic Optimization.
- Earlier in my Ph.D., I focused on 3D scene understanding. In particular, we built a self-supervised representation of 3D scenes to better retrieve similar 3D subscenes. There are two motivations behind our work: i) We can compress a large database of 3D scenes by deduplicating similar 3D subscenes or alternatively ii) We can edit the similar 3D subscenes to generate new 3D scenes and enrich the database.