Hello! Welcome to Lewis Liu’s homepage :)
I am currently a master student at Mila and DIRO in Montreal, supervised by Prof. Simon Lacoste-Julien. I also closely work with Damien Scieur at Samsung SAIL Montreal, as well as amazing researchers from University of Montreal, McGill University, Northwestern University, MIT and IBM Research.
In general, I am broadly interested in all aspects of optimization, design and analysis of algorithms, probability, and statistics. Specifically, my research focuses on the application of techniques from pure mathematics to the solutions of fundamental problems in optimization, statistical inference, and algorithm design, which leads to two main goals of my research. One is to develop both computationally and statistically efficient algorithms for modern optimization and control problems with provable guarantees while understanding both asymptotic and non-asymptotic behaviors of existing models (e.g., random walks). The other is devoted to designing efficient approximation algorithms for discrete optimization with massive data while understanding the computational complexity and limits of existing methods. In addition, applications of the theory to practical operation research problems (e.g., large-scale machine learning, transportation, supply chains, etc.) are also motivating my research.
- Convex & Nonconvex Optimization
- Probability Theory, Markov Chains and Random Walks
- Design and Analysis of Algorithms, Combinatorics
- Nonparametric Statistics
- Theory of Control and Reinforcement Learning
In my spare time, I enjoy competitive programming and chess, where the Fried Liver Attack is one of my favorite opening tricks due to its name; See also its variant Lolli Attack with a similarly lovely name. I am also a fan of anime and cool robotics hacking. I used to collect amazing stamps.