Adversarial robustness of deep reinforcement learning

• 主要作者: Qu, Xinghua
• 其他作者: Ong Yew Soon
• 格式: Thesis-Doctor of Philosophy
• 語言: English
• 出版: Nanyang Technological University 2022
• 主題: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• 在線閱讀: https://hdl.handle.net/10356/154587

Over the past decades, the advancements in deep reinforcement learning (DRL) have demonstrated that deep neural network (DNN) policies can be trained to prescribe near-optimal actions in many complex tasks. Unfortunately, DNN policies are shown to be vulnerable to adversarial perturbations in the in...