Victor Liao

I am a senior at the University of Waterloo studying Pure Mathematics, Combinatorics & Optimization, and Computer Science.

My main research interests are at the intersection of optimization and math with applications to data science. I am broadly interested in non-convex optimization, structure in optimization, and the application of mathematical theory from areas like non-smooth/variational analysis, semialgebraic/differential geometry, and differential inclusions to optimization. I am also interested in mathematical analysis with notable topics being functional analysis, Banach algebras and operator theory.

I work with Nicolas Boumal at École Polytechnique Fédérale de Lausanne on optimization and Manopt. I also work with Yaoliang Yu at the University of Waterloo on optimization and machine learning.

Previously, I spent time collaborating with Amir-massoud Farahmand at the Vector Institute and Martha White at the University of Alberta on optimization and reinforcement learning.

Email  /  Google Scholar  /  Github

profile photo
Research Papers
(Reverse Chronological Order)
Robust Losses for Learning Value Functions
Andrew Patterson, Victor Liao, Martha White
To Appear in: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022.   (Impact Factor: 24.31)

Value Gradient weighted Model-Based Reinforcement Learning
Claas Voelcker, Victor Liao, Animesh Garg, Amir-massoud Farahmand
International Conference on Learning Representations (ICLR), 2022.   (Spotlight Presentation rate: 5.3%)


Coming Soon
Victor Liao, Zeou Hu, Yaoliang Yu
In preparation, 2023

Last updated: January 15, 2023 Page template from here!