Zijian Huang is a Master of Science in Computer Science student at University of Illinois at Urbana-Champaign(UIUC), supervised by Prof. Bo Li. His research interests include Machine Learning, Security and Computer Vision. Specifically, he is interested in the robustness of machine learning models. He is generally interested in the theoretical part of adversarial machine learning, like the certified robustness of reinforcement learning algorihtms, and also the pratical part, such as robustness of object detection models. Also, he has done some CV research projects, including 3D human pose detection and image/video synthesis.
Download my resumé.
Master of Science in Computer Science, 2021 - present
University of Illinois at Urbana-Champaign(UIUC)
BSc in Artificial Intelligence, 2016 - 2020
The Hong Kong University of Science and Technology
100%
80%
60%
60%
60%
We propose a framework for certifying the robustness of Q-learning algorithms, which is notably the first that provides the robustness certification w.r.t. the cumulative reward; We propose two robustness certification criteria for Q-learning algorithms, together with corre- sponding certification algorithms based on global and local smoothing strategies; We theoretically prove the certification radius for input state and lower bound of perturbed cumula- tive reward under bounded adversarial state perturbations; We conduct extensive experiments to provide certification for nine empirically robust RL algorithms on multiple RL environments. We provide several interesting observations which would further inspire the development of robust RL algorithms.