arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.07430cs.AIcs.LG
收藏

基于学习模型的搜索和规划的连续控制

Continuous Control for Searching and Planning with a Learned Model

Iowa State University(爱荷华州立大学);Ecole Centrale de Nantes(南特中央理工学院)

Xuxi Yang, Werner Duvaud, Peng Wei

摘要:

具有规划能力的决策代理在国际象棋、将棋和围棋等具有挑战性的领域取得了巨大的成功。为了将规划能力推广到更一般的任务,其中环境动态对代理不可用,研究人员提出了MuZero算法,该算法可以通过与环境的交互来学习动态模型。在本文中,我们提供了一种方法和必要的理论结果,以扩大MuZero算法更广义的环境与连续的行动空间。通过在两个相对低维的MuJoCo环境中的数值结果,我们表明该算法优于软演员-评论家(SAC)算法,这是一种最先进的无模型深度强化学习算法。

英文摘要:

Decision-making agents with planning capabilities have achieved huge success in the challenging domain like Chess, Shogi, and Go. In an effort to generalize the planning ability to the more general tasks where the environment dynamics are not available to the agent, researchers proposed the MuZero algorithm that can learn the dynamical model through the interactions with the environment. In this paper, we provide a way and the necessary theoretical results to extend the MuZero algorithm to more generalized environments with continuous action space. Through numerical results on two relatively low-dimensional MuJoCo environments, we show the proposed algorithm outperforms the soft actor-critic (SAC) algorithm, a state-of-the-art model-free deep reinforcement learning algorithm.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑