Knowledge Transfer for Deep Reinforcement Learning with Hierarchical Experience Replay

• المؤلفون الرئيسيون: Yin, Haiyan, Pan, Sinno Jialin
• مؤلفون آخرون: School of Computer Science and Engineering
• التنسيق: Conference or Workshop Item
• اللغة: English
• منشور في: 2017
• الموضوعات: Deep reinforcement learning; Policy distillation
• الوصول للمادة أونلاين: https://hdl.handle.net/10356/83043; http://hdl.handle.net/10220/42453; https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14478
• المؤسسة: Nanyang Technological University

The process for transferring knowledge of multiple reinforcement learning policies into a single multi-task policy via distillation technique is known as policy distillation. When policy distillation is under a deep reinforcement learning setting, due to the giant parameter size and the huge state s...