Search results
Feb 11, 2021 · Robust Policy Gradient against Strong Data Corruption. Xuezhou Zhang, Yiding Chen, Xiaojin Zhu, Wen Sun. We study the problem of robust reinforcement learning under adversarial corruption on both rewards and transitions. Our attack model assumes an \textit {adaptive} adversary who can arbitrarily corrupt the reward and transition at every step ...
Dec 6, 2019 · View a PDF of the paper titled Hyperbolic Graph Attention Network, by Yiding Zhang and 4 other authors View PDF Abstract: Graph neural network (GNN) has shown superior performance in dealing with graphs, which has attracted considerable research attention recently.
Yiding Zhang. Lyujie Chen. Aiping Lu. Inference of disease-gene associations helps unravel the pathogenesis of diseases and contributes to the treatment. Although many machine learning-based ...
Dec 6, 2020 · Yiding Liu [email protected] College of Chemistry and Chemical Engineering, Southwest Petroleum University, Chengdu, 610500 P. R. China. E-mail: [email protected]; [email protected]; [email protected]; [email protected] Search for more papers by this author
Dec 6, 2019 · Download a PDF of the paper titled Hyperbolic Graph Attention Network, by Yiding Zhang and 4 other authors Download PDF Abstract: Graph neural network (GNN) has shown superior performance in dealing with graphs, which has attracted considerable research attention recently.
Wang, Zhang, and Shi 2019). One key property of hyperbolic spaces is that they ex-pand faster than Euclidean spaces, because Euclidean spaces expand polynomially while hyperbolic spaces expand ex-ponentially. For instance, each tile in Fig. 1(a) is of equal area in hyperbolic space but diminishes towards zero in Eu-clidean space towards the ...
Xuezhou Zhang 1Yiding Chen Jerry Zhu Wen Sun2 Abstract We study the problem of robust reinforcement learning under adversarial corruption on both re-wards and transitions. Our attack model assumes an adaptive adversary who can arbitrarily corrupt the reward and transition at every step within an episode, for at most "-fraction of the learning ...
