提出三种用于随机最短路径问题的收敛强化学习算法，性能优于现有方法。
Convergent Reinforcement Learning Algorithms for Stochastic Shortest Path Problem
- 基于表格和函数逼近两种设置设计新算法，保证几乎必然收敛。
- 表格算法在多个任务中表现超越经典收敛算法，提升明显。
- 函数逼近算法在复杂场景下稳定可靠，适合实际应用部署。
本文针对随机最短路径（SSP）问题，在表格设定下提出了两种算法，并在函数逼近设定下提出一种算法。SSP问题是强化学习中的重要类别，其他类型的代价准则均可转化为该框架。我们证明了所有算法的渐近几乎必然收敛性。实验表明，所提表格算法相比其他已知收敛算法表现出更优性能；函数逼近算法在相应设置中也展现出可靠性能，优于现有方法。
原文摘要 · Abstract (English)
In this paper we propose two algorithms in the tabular setting and an algorithm for the function approximation setting for the Stochastic Shortest Path (SSP) problem. SSP problems form an important class of problems in Reinforcement Learning (RL), as other types of cost-criteria in RL can be formulated in the setting of SSP. We show asymptotic almost-sure convergence for all our algorithms. We observe superior performance of our tabular algorithms compared to other well-known convergent RL algorithms. We further observe reliable performance of our function approximation algorithm compared to other algorithms in the function approximation setting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。