arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2209.09058cs.LGcs.AI
收藏

强化学习中干预鲁棒性的度量

Measuring Interventional Robustness in Reinforcement Learning

Knowledge Discovery Laboratory(知识发现实验室);College of Information and Computer Sciences(信息与计算机科学学院);University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校)

Katherine Avery, Jack Kenney, Pracheta Amaranath, Erica Cai, David Jensen

摘要:

强化学习的最新工作集中在学习策略的几个特征上,这些特征超越了最大化奖励。这些属性包括公平性、可解释性、泛化性和鲁棒性。在本文中,我们定义了干预鲁棒性(IR),有多少可变性引入到学习政策的附带方面的培训过程中,如训练数据的顺序或代理采取的特定探索行动的措施。当它产生的代理在干预下采取非常相似的行动时,训练过程具有高IR,尽管训练过程的这些附带方面存在变化。我们开发了一种直观的IR定量测量方法,并在三个Atari环境中的八种算法中计算了数十种干预措施和状态。从这些实验中,我们发现IR随训练量和算法类型而变化,并且高性能并不意味着高IR,正如人们所期望的那样。

英文摘要:

Recent work in reinforcement learning has focused on several characteristics of learned policies that go beyond maximizing reward. These properties include fairness, explainability, generalization, and robustness. In this paper, we define interventional robustness (IR), a measure of how much variability is introduced into learned policies by incidental aspects of the training procedure, such as the order of training data or the particular exploratory actions taken by agents. A training procedure has high IR when the agents it produces take very similar actions under intervention, despite variation in these incidental aspects of the training procedure. We develop an intuitive, quantitative measure of IR and calculate it for eight algorithms in three Atari environments across dozens of interventions and states. From these experiments, we find that IR varies with the amount of training and type of algorithm and that high performance does not imply high IR, as one might expect.

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。

↑