I am a first-year PhD student in computer science at the University of Toronto, supervised by Prof. Chris Maddison. Previously, I completed my MMath and BMath in computer science at the University of Waterloo, supervised by Prof. Pascal Poupart.
I am interested in improving the reasoning and inference abilities of intelligence agents. Some of my past work focused on designing self-supervised pre-training paradigms for reasoning, developing exact inference models with tractability guarantee and improving combinatorial solvers using Bayesian inference. I am also interested in representation learning and uncertainty quantification for neural networks.
* below indicates equal contribution
Augment with Care: Contrastive Learning for Combinatorial Problems
Haonan Duan*, Pashootan Vaezipoor*, Max B. Paulus, Yangjun Ruan, Chris J. Maddison
International Conference on Machine Learning (ICML), 2022.
Distributional Reinforcement Learning with Monotonic Splines
Yudong Luo, Guiliang Liu, Haonan Duan, Oliver Schulte, Pascal Poupart
International Conference on Learning Representations (ICLR), 2022
Multiple Moment Matching Inference: A Flexible Approximate Inference Algorithm
Haonan Duan, Pascal Poupart
ICML 2021 Workshop on Uncertainty & Robustness in Deep Learning
Online Bayesian Moment Matching based SAT Solver Heuristics
Haonan Duan*, Saeed Nejati*, George Trimponias, Pascal Poupart and Vijay Ganesh
International Conference on Machine Learning (ICML), 2020.
Discriminative Training of Feed-Forward and Recurrent Sum-Product Networks by Extended Baum-Welch
Haonan Duan, Abdullah Rashwan, Pascal Poupart and Zhitang Chen
International Journal of Approximate Reasoning (IJAR), Volume 124, September 2020, Pages 66-81
Conference reviewer: ICLR 2021, ICML 2022, NeurIPS 2022
Workshop reviewer: AAAI 2022 workshop on Machine Learning for Operation Research
Software Engineer Intern @
Jan. 2021 - Apr. 2021
Data Scientist Intern @
May. 2020 - Aug. 2021
Selected Honors & Awards
I am grateful for the organizations and people below to support my study and research: