About me
I am an assistant professor in the Computer Science Department at University of Houston. Prior to that, I was a Postdoctoral Scholar in AI-EDGE Institute at The Ohio State University, advised by Prof. Ness Shroff, Prof. Yingbin Liang, and Prof. Anish Arora. I received my B.Eng. degree in Electrical Engineering from Zhejiang University, Hangzhou, China, in 2013, my M.S. degree in Telecommunications from The Hong Kong University of Science and Technology, Hong Kong, in 2014, and my Ph.D. degree in Electrical Engineering from Arizona State University, in 2021, under the co-supervision of Prof. Junshan Zhang and Prof. Lei Ying.
News
[2025.03] Check our recent survey paper on Mixture-of-Experts here, which provides a comprehensive review of its basics, algorithm design in multiple learning paradigms, theory and applications in CV/NLP.
[2025.02] Our book “Continual and Reinforcement Learning for Edge AI” has been published by Springer and will be available soon.
[2025.02] One paper on continual learning has been accepted by ISQED 2025 and one paper on bilevel optimization has been accepted by CPAL 2025. Our paper on the theory of MoE in continual learning has also been selected as a Spotlight (5.1%) in ICLR 2025.
[2025.01] Our paper on the theory of MoE in continual learning has been accepted by ICLR 2025, which provides the first comprehensive theoretical understanding of how MoE works in continual learning.
[2024.12] I will serve as a TPC member of MobiHoc 2025 and ITW 2025.
[2024.10] I will serve as a guest editor for IEEE/ACM Transactions on Networking Special Issue on AI and Networking, which welcomes cutting-edge research findings in the areas of “AI for networks” and “AI on networks”. Check the link here if you are interested.
[2024.05] I will serve as a TPC member of INFOCOM 2025.
[2024.05] Two papers on offline and offline-to-online imitation learning have been accepted by ICML 2024. Congratulations to Sheng.
[2024.01] Our paper on adversarial training has been accepted by ICLR 2024. We propose a novel doubly-robust instance-reweighed adversarial training method based on bilevel optimization and distributionally robust optimization, which significantly boosts the robustness on the most vulnerable examples.
[2023.12] I have accepted the invitation to serve as a Program Committee member of ICDCS 2024.
[2023.11] One paper on online meta-learning has been accepted by CPAL 2024.
[2023.09] Our paper on online bilevel optimization has been accepted by NeurIPS 2023.
[2023.09] I have accepted the invitation to serve as a Program Committee member of SDM 2024.
[2023.08] Our paper “Scheduling Real-time Wireless Traffic: A Network-aided Offline Reinforcement Learning Approach” has been accepted by IEEE Internet of Things Journal.
[2023.04] Our paper “Theory on Forgetting and Generalization of Continual Learning” is accepted by ICML 2023. Congratulations to Peizhong and other authors.
[2023.04] Our paper “Warm-Start Actor-Critic: From Approximation Error to Sub-optimality Gap” is accepted by ICML 2023 as an oral presentation. Congratulations to Hang and other authors.