摘要
综述了再励学习(Reinforcement Learning)的原理,主要算法,基于神经网络的实现及其在智能控制中的作用,探讨了应进一步研究的问题.
Principle,typical algorithm,neural network-based implementation of reinforcement learning, as well as its function in intelligent control are overviewed in this paper.Some problems for further study are also discussed.
出处
《信息与控制》
CSCD
北大核心
1996年第1期28-34,共7页
Information and Control
基金
攀登计划认知科学(神经网络)重大关键项目资助课题
关键词
再励学习
学习控制
智能控制
reinforcement learning
learning control
intelligent control
