Realistic Actor-Critic: A framework for balance between value overestimation and underestimation

IntroductionThe value approximation bias is known to lead to suboptimal policies or catastrophic overestimation bias accumulation that prevent the agent from making the right decisions between exploration and exploitation. Algorithms have been proposed to mitigate the above contradiction. However, w...

Full description

Bibliographic Details

Published in:	Frontiers in Neurorobotics
Main Authors:	Sicen Li, Qinyun Tang, Yiming Pang, Xinmeng Ma, Gang Wang
Format:	Article
Language:	English
Published:	Frontiers Media S.A. 2023-01-01
Subjects:	reinforcement learning (RL) robot control estimation bias exploration-exploitation dilemma uncertainty
Online Access:	https://www.frontiersin.org/articles/10.3389/fnbot.2022.1081242/full