Hoeffding和Bernstein的策略选择竞赛Hoeffding and Bernstein Races for Selecting Policies in Evolutionary Direct Policy Search |
|
课程网址: | https://videolectures.net/videos/icml09_igel_hbrs |
主讲教师: | Christian Igel |
开课单位: | 会议 |
开课时间: | 2009-08-26 |
课程语种: | 英语 |
中文简介: | 在从各种来源进行强化学习时会出现不确定性,因此有必要考虑基于几次推出的统计数据来评估行为策略。我们在CMA-ES中添加了基于Hoeffding和经验Bernstein竞争的自适应不确定性处理,这是一种为直接策略搜索提出的可变度量进化策略。不确定性处理单独调整评估政策时考虑的事件数量。性能估计保持足够准确,以便对候选策略进行足够好的排名,这反过来又足以让CMA-ES找到更好的解决方案。这提高了算法的学习速度和鲁棒性。 |
课程简介: | Uncertainty arises in reinforcement learning from various sources, and therefore it is necessary to consider statistics based on several roll-outs for evaluating behavioral policies. We add an adaptive uncertainty handling based on Hoeffding and empirical Bernstein races to the CMA-ES, a variable metric evolution strategy proposed for direct policy search. The uncertainty handling adjusts individually the number of episodes considered for the evaluation of a policy. The performance estimation is kept just accurate enough for a sufficiently good ranking of candidate policies, which is in turn sufficient for the CMA-ES to find better solutions. This increases the learning speed as well as the robustness of the algorithm. |
关 键 词: | 直接策略; 策略选择; 强化学习 |
课程来源: | 视频讲座网 |
数据采集: | 2025-04-25:liyq |
最后编审: | 2025-04-25:liyq |
阅读次数: | 5 |