I am a PhD student at MILA and the Université de Montréal within the DIRO department. I am supervised by Ioannis Mitliagkas. My current research interests include (but are not limited to), online learning, optimization, reinforcement learning and algorithmic game theory. Previously, I completed my masters in statistical machine learning at the University of Alberta under the co-supervision of James Wright and Matthew Taylor where I worked on combining function approximation and online learning for saddle point computation. Before attending the University of Alberta I completed my bachelor of science in actuarial mathematics at Concordia University and worked as an actuarial analyst for several insurance companies.

Contact: ryan.dorazio [at] mila.quebec

Publications

Stochastic mirror descent: Convergence analysis and adaptive variants via the mirror stochastic Polyak stepsize
Ryan D’Orazio, Nicolas Loizou, Issam Laradji, Ioannis Mitliagkas
TMLR 2023
Paper

Abstracting Imperfect Information Away from Two-Player Zero-Sum Games
Samuel Sokota, Ryan D’Orazio, Chun Kai Ling, David J Wu, J Zico Kolter, Noam Brown
ICML 2023
Paper

A Unified Approach to Reinforcement Learning, Quantal Response Equilibria, and Two-Player Zero-Sum Games
Sam Sokota*, Ryan D’Orazio*, Zico Kolter, Nicolas Loizou, Marc Lanctot, Ioannis Mitliagkas, Noam Brown, Christian Kroer
ICLR 2023
Paper

Efficient Deviation Types and Learning for Hindsight Rationality in Extensive-Form Games
Dustin Morrill, Ryan D’Orazio, Marc Lanctot, James R. Wright, Amy Greenwald, Michael Bowling
ICML 2021
Paper

Hindsight and Sequential Rationality of Correlated Play
Dustin Morrill, Ryan D’Orazio, Reca Sarfati, Marc Lanctot, James R. Wright, Amy Greenwald, Michael Bowling
AAAI 2021
Paper

Solving Common-Payoff Games with Approximate Policy Iteration
Samuel Sokota*, Edward Lockhart*, Finbarr Timbers, Elnaz Davoodi, Ryan D’Orazio, Neil Burch, Martin Schmid, Michael Bowling, Marc Lanctot
AAAI 2021
Paper / Code

Alternative Function Approximation Parameterizations for Solving Games: An Analysis of f-Regression Counterfactual Regret Minimization
Ryan D’Orazio*, Dustin Morrill*, James Wright, Michael Bowling
AAMAS 2020
Paper

Simultaneous Prediction Intervals for Patient-Specific Survival Curves
Samuel Sokota*, Ryan D’Orazio*, Khurram Javed, Humza Haider, Russell Greiner
IJCAI 2019
Paper / Code

Workshop Papers

On Stochastic Mirror Descent: Convergence Analysis and Adaptive Variants
Ryan D’Orazio, Nicolas Loizou, Issam Laradji, Ioannis Mitliagkas
Beyond first-order methods in ML systems, ICML 2021
Full Arxiv Version / Workshop Version / Code

Optimistic and Adaptive Lagrangian Hedging
Ryan D’Orazio, Ruitong Huang
Workshop on Reinforcement Learning and Games, AAAI 2021
Paper / Poster

Bounds for Approximate Regret-Matching Algorithms
Ryan D’Orazio, Dustin Morrill, James R. Wright
Smooth Games Optimization and Machine Learning Workshop, NeurIPS 2019
Paper

Thesis

Regret Minimization with Function Approximation in Extensive-Form Games
MSc Outstanding Thesis Award (Dept of Computing Science)
Ryan D’Orazio
Masters Thesis
University of Alberta
Paper