arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1911.04448cs.LGstat.ML
收藏

实时的强化学习

Real-Time Reinforcement Learning

Mila(米拉研究所);Element AI(Element AI公司);Université de Montréal(蒙特利尔大学);Polytechnique Montréal(蒙特利尔综合理工学院)

Simon Ramstedt, Christopher Pal

摘要:

马尔可夫决策过程(MDP)是强化学习(RL)中大多数算法的数学框架,通常被错误地假设代理环境的状态在动作选择期间不会改变。随着基于MDP的RL系统开始在现实世界的安全关键情况中找到应用,经典MDP的假设与实时的【形】计算的现实之间的这种不匹配可能导致不期望的结果。在本文中,我们引入了一个新的框架,在其中状态和动作同时演变,并显示它是如何与经典的MDP制定。我们分析现有的算法下的新的实时的【形】配方,并显示为什么他们是次优的实时的【形】使用时。然后,我们使用这些见解来创建一个新的算法实时的【形】Actor-Critic(RTAC),它在实时的【形】和非实时设置中的性能都优于现有的最先进的连续控制算法Soft Actor-Critic。代码和视频可以在https://github.com/rmst/rtrl上找到。

英文摘要:

Markov Decision Processes (MDPs), the mathematical framework underlying most algorithms in Reinforcement Learning (RL), are often used in a way that wrongfully assumes that the state of an agent's environment does not change during action selection. As RL systems based on MDPs begin to find application in real-world safety critical situations, this mismatch between the assumptions underlying classical MDPs and the reality of real-time computation may lead to undesirable outcomes. In this paper, we introduce a new framework, in which states and actions evolve simultaneously and show how it is related to the classical MDP formulation. We analyze existing algorithms under the new real-time formulation and show why they are suboptimal when used in real-time. We then use those insights to create a new algorithm Real-Time Actor-Critic (RTAC) that outperforms the existing state-of-the-art continuous control algorithm Soft Actor-Critic both in real-time and non-real-time settings. Code and videos can be found at https://github.com/rmst/rtrl.

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。

↑