arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2311.05511cs.LGcs.AIcs.DS
收藏

任意时刻约束强化学习

Anytime-Constrained Reinforcement Learning

University of Wisconsin-Madison(威斯康星大学麦迪逊分校)

Jeremy McMahan, Xiaojin Zhu

AI总结:

本文研究带任意时刻约束的 cMDP,通过归约到无约束 MDP 实现高效规划与学习,并针对 NP 难情形提出可证明的近似算法,达到最坏情况下的最优近似保证。

摘要:

我们引入并研究了带任意时刻约束的约束马尔可夫决策过程(cMDPs)。任意时刻约束要求智能体在任何时间点都几乎必然不违反其预算。尽管马尔可夫策略已不再充分,我们证明存在由累积代价增强的最优确定性策略。事实上,我们给出了从任意时刻约束 cMDP 到无约束 MDP 的固定参数可处理归约。只要代价的精度是 cMDP 规模的对数量级,我们的归约就能为表格型 cMDP 提供时间和样本高效的计划与学习算法。然而,我们也证明,一般而言计算非平凡的近似最优策略是 NP 难的。为绕过这一瓶颈,我们设计了可证明的近似算法,只要最大支撑代价受 cMDP 或绝对预算的多项式约束,这些算法就能高效地计算或学习任意精度的近似可行策略,并具有最优值。鉴于我们的困难性结果,在最坏情况分析下,我们的近似保证是最好的可能结果。

英文摘要:

We introduce and study constrained Markov Decision Processes (cMDPs) with anytime constraints. An anytime constraint requires the agent to never violate its budget at any point in time, almost surely. Although Markovian policies are no longer sufficient, we show that there exist optimal deterministic policies augmented with cumulative costs. In fact, we present a fixed-parameter tractable reduction from anytime-constrained cMDPs to unconstrained MDPs. Our reduction yields planning and learning algorithms that are time and sample-efficient for tabular cMDPs so long as the precision of the costs is logarithmic in the size of the cMDP. However, we also show that computing non-trivial approximately optimal policies is NP-hard in general. To circumvent this bottleneck, we design provable approximation algorithms that efficiently compute or learn an arbitrarily accurate approximately feasible policy with optimal value so long as the maximum supported cost is bounded by a polynomial in the cMDP or the absolute budget. Given our hardness results, our approximation guarantees are the best possible under worst-case analysis.

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。

↑