2021/02/27 by Tadashi Kozuno, Yunhao Tang, Kozuno, Tadashi +13 · 1 citation
Computer Science · #Artificial Intelligence (cs.AI) #Evolutionary Algorithms and Applications #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Optimization and Search Problems #Reinforcement Learning in Robotics
paper · pdf · doi:10.48550/arxiv.2103.00107
openalex publication_date 2021/02/27 · openalex created_date 2021/08/02 · openalex updated_date 2026/07/28
Off-policy multi-step reinforcement learning algorithms consist of conservative and non-conservative algorithms: the former actively cut traces, whereas the latter do not. Recently, Munos et al. (2016) proved the convergence of conservative algorithms to an optimal Q-function. In contrast, non-conservative algorithms are thought to be unsafe and have a limited or no theoretical guarantee. Nonetheless, recent studies have shown that non-conservative algorithms empirically outperform conservative ones. Motivated by the empirical results and the lack of theory, we carry out theoretical analyses of Peng's Q(λ), a representative example of non-conservative algorithms. We prove that it also converges to an optimal policy provided that the behavior policy slowly tracks a greedy policy in a way similar to conservative policy iteration. Such a result has been conjectured to be true but has not been proven. We also experiment with Peng's Q(λ) in complex continuous control tasks, confirming that Peng's Q(λ) often outperforms conservative algorithms despite its simplicity. These results indicate that Peng's Q(λ), which was thought to be unsafe, is a theoretically-sound and practically effective algorithm.