arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2306.03072cs.LG
收藏

零样本强化学习中的探索以泛化

Explore to Generalize in Zero-Shot RL

Technion – Israel Institute of Technology(以色列理工学院)

Ev Zisselman, Itai Lavie, Daniel Soudry, Aviv Tamar

AI总结:

针对零样本强化学习中的过拟合问题,提出Explore to Generalize算法,通过训练奖励优化智能体集成并在测试时结合探索性动作实现泛化,在ProcGen Maze和Heist等任务上取得最先进结果。

摘要:

我们研究强化学习中的零样本泛化问题——在一组训练任务上优化策略,以在相似但未见过的测试任务上表现良好。为了缓解过拟合,先前的工作探索了不同的任务不变性概念。然而,在诸如ProcGen Maze等问题上,不存在对任务可视化保持不变性的充分解,因此基于不变性的方法会失败。我们的见解是,学习一个能有效探索领域环境的策略,比学习一个针对特定任务最大化奖励的策略更难被记忆,因此我们预期这种习得行为能够良好泛化;我们确实在几个对基于不变性方法而言困难的领域上通过实验证明了这一点。我们的Explore to Generalize算法(ExpGen)建立在这一见解之上:我们额外训练一组优化奖励的智能体集成。在测试时,要么该集成对某个动作达成一致,此时我们能够良好泛化;要么我们采取探索性动作,这些动作能够良好泛化,并将我们带到状态空间的新区域,在那里该集成可能再次达成一致。我们表明,我们的方法在ProcGen挑战中迄今未能实现有效泛化的任务上达到了最先进水平,在200个训练关卡下,Maze任务的成功率达到83%,Heist任务的成功率达到74%。ExpGen还可以与基于不变性的方法结合,以兼得两者之长,在ProcGen上创造了新的最先进结果。

英文摘要:

We study zero-shot generalization in reinforcement learning-optimizing a policy on a set of training tasks to perform well on a similar but unseen test task. To mitigate overfitting, previous work explored different notions of invariance to the task. However, on problems such as the ProcGen Maze, an adequate solution that is invariant to the task visualization does not exist, and therefore invariance-based approaches fail. Our insight is that learning a policy that effectively $\textit{explores}$ the domain is harder to memorize than a policy that maximizes reward for a specific task, and therefore we expect such learned behavior to generalize well; we indeed demonstrate this empirically on several domains that are difficult for invariance-based approaches. Our $\textit{Explore to Generalize}$ algorithm (ExpGen) builds on this insight: we train an additional ensemble of agents that optimize reward. At test time, either the ensemble agrees on an action, and we generalize well, or we take exploratory actions, which generalize well and drive us to a novel part of the state space, where the ensemble may potentially agree again. We show that our approach is the state-of-the-art on tasks of the ProcGen challenge that have thus far eluded effective generalization, yielding a success rate of $83\%$ on the Maze task and $74\%$ on Heist with $200$ training levels. ExpGen can also be combined with an invariance based approach to gain the best of both worlds, setting new state-of-the-art results on ProcGen.

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。

↑