arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1402.0635stat.MLcs.AIcs.LGcs.SYeess.SY
收藏

通过随机价值函数实现泛化与探索

Generalization and Exploration via Randomized Value Functions

Stanford University(斯坦福大学)

Ian Osband, Benjamin Van Roy, Zheng Wen

AI总结:

本文提出随机最小二乘价值迭代算法(RLSVI),通过线性参数化价值函数实现高效的探索与泛化,证明其在无先验知识学习中的近优性能。

摘要:

我们提出了随机最小二乘价值迭代(RLSVI)——一种新的强化学习算法,旨在通过线性参数化价值函数实现高效的探索与泛化。我们解释了使用玻尔兹曼或epsilon-贪婪探索的最小二乘价值迭代版本为何效率低下,并通过计算结果展示了RLSVI带来的显著效率提升。进一步,我们建立了RLSVI预期遗憾的上界,证明其在无先验知识学习中的近最优性。更广泛地说,我们的结果表明,随机价值函数为解决强化学习中的关键挑战——合成高效探索与有效泛化——提供了一种有前景的方法。

英文摘要:

We propose randomized least-squares value iteration (RLSVI) -- a new reinforcement learning algorithm designed to explore and generalize efficiently via linearly parameterized value functions. We explain why versions of least-squares value iteration that use Boltzmann or epsilon-greedy exploration can be highly inefficient, and we present computational results that demonstrate dramatic efficiency gains enjoyed by RLSVI. Further, we establish an upper bound on the expected regret of RLSVI that demonstrates near-optimality in a tabula rasa learning context. More broadly, our results suggest that randomized value functions offer a promising approach to tackling a critical challenge in reinforcement learning: synthesizing efficient exploration and effective generalization.

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。

↑