arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.10888cs.LGcs.AIstat.ML
收藏

风险规避强化学习的均值-方差策略迭代

Mean-Variance Policy Iteration for Risk-Averse Reinforcement Learning

University of Oxford(牛津大学);Auburn University(奥本大学)

Shangtong Zhang, Bo Liu, Shimon Whiteson

摘要:

本文提出了一个均值-方差策略迭代(MVPI)框架,用于贴现无限时域MDP中的风险厌恶控制,该框架优化了每步奖励随机变量的方差。MVPI具有很大的灵活性,因为任何策略评估方法和风险中性控制方法都可以用于现成的风险规避控制,无论是在策略上还是在策略外。这种灵活性减少了风险中性控制和风险厌恶控制之间的差距,并通过直接在一个新的增强MDP上工作来实现。我们提出了风险厌恶TD 3作为MVPI实例化的一个例子,它优于香草TD 3和许多以前的风险厌恶控制方法在具有风险意识的性能指标下挑战Mujoco机器人仿真任务。这种风险厌恶的TD 3是第一个将确定性策略和离策略学习引入风险厌恶的强化学习的,这两者都是我们在Mujoco领域中展示的性能提升的关键。

英文摘要:

We present a mean-variance policy iteration (MVPI) framework for risk-averse control in a discounted infinite horizon MDP optimizing the variance of a per-step reward random variable. MVPI enjoys great flexibility in that any policy evaluation method and risk-neutral control method can be dropped in for risk-averse control off the shelf, in both on- and off-policy settings. This flexibility reduces the gap between risk-neutral control and risk-averse control and is achieved by working on a novel augmented MDP directly. We propose risk-averse TD3 as an example instantiating MVPI, which outperforms vanilla TD3 and many previous risk-averse control methods in challenging Mujoco robot simulation tasks under a risk-aware performance metric. This risk-averse TD3 is the first to introduce deterministic policies and off-policy learning into risk-averse reinforcement learning, both of which are key to the performance boost we show in Mujoco domains.

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。

↑