.png){kind=icon}
返回
Distributed Policy Evaluation Under Multiple Behavior Strategies
.png){kind=icon}
OA
AI
DOI：10.1109/TAC.2014.2368731.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
We apply diffusion strategies to develop a fully-distributed cooperative reinforcement learning algorithm in which agents in a network communicate only with their immediate neighbors to improve predictions about their environment. The algorithm can also be applied to off-policy learning, meaning that the agents can predict the response to a behavior different from the actual policies they are following. The proposed distributed strategy is efficient, with linear complexity in both computation time and memory footprint. We provide a mean-square-error performance analysis and establish convergence under constant stepsize updates, which endow the network with continuous learning capabilities. The results show a clear gain from cooperation: when the individual agents can estimate the solution, cooperation increases stability and reduces bias and variance of the prediction error; but, more importantly, the network is able to approach the optimal solution even when none of the individual agents can (e.g., when the individual behavior policies restrict each agent to sample a small portion of the state space).
Keyword:
Adaptive networks
Arrow-Hurwicz algorithm
diffusion strategies
distributed processing
gradient temporal difference
mean-square-error
reinforcement learning
saddle-point problem
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
7
论文数:
1.3W
被引数:
6.7W
机构
引用论文
Critical Evaluation and Thermodynamic Optimization of the Li2O-Al2O3 and Li2O-MgO-Al2O3 Systems
.png){kind=icon}
PREAI
分享收藏
查看更多内容