一种激励学习遗忘算法

A Kind of Forgetting Reinforcement Learning Algorithm

提出了一种基于动态规划方法的激励学习遗忘算法,这个算法是通过将记忆心理学中有关遗忘的基本原理引入到值函数的激励学习中,导出了一类用动态规划方法解决激励学习问题的比较好的方法,即Forget-DP算法。

In this paper we give a forgetting algorithm,which introduces the forgetting principles in psychology to reinforcement learning about value functions.Using forgetting algorithm,we discuss that problem describing above.