| 基于RBF函数状态离散化的激励学习 |
| |
| 引用本文： | 田建军,唐中勇.基于RBF函数状态离散化的激励学习[J].太原师范学院学报(自然科学版),2006,5(3):50-53. |
| |
| 作者姓名： | 田建军 唐中勇 |
| |
| 作者单位： | 1. 湖南公安高等专科学校,计算机系,湖南,长沙,410138
2. 长沙理工大学,计算机通讯工程学院,湖南,长沙,410076 |
| |
| 摘 要： | 介绍了激励学习和两类学习算法：Q学习和SARSA学习，提出一类基于RBF函数的特征状态离散化方法，并对该方法进行了初步的实验比较．
|
| 关 键 词： | 激励学习 特征状态 状态离散化 RBF函数 |
| 文章编号： | 1672-2027（2006）03-0050-04 |
| 收稿时间： | 2006-03-09 |
| 修稿时间： | 2006年3月9日 |
The State Discretization Based on RBF Function for the Reinforcement Learning |
| |
| Tian Jianjun,Tang Zhongyong.The State Discretization Based on RBF Function for the Reinforcement Learning[J].Journal of Taiyuan Normal University:Natural Science Edition,2006,5(3):50-53. |
| |
| Authors: | Tian Jianjun Tang Zhongyong |
| |
| Institution: | 1. Department of Computer Science,Hunan Public Security College,Changsha 410138; 2. Department of Computer and Communication,Changsha University of Science and Technology,Changsha 410076,China |
| |
| Abstract: | Reinforcement Learning and two classes of learning algorithms is introduced. A class of the state discretization based on RBF function for the Reinforcement Learning is proposed and preliminary empirical results are presented to compare the performance of the new method. |
| |
| Keywords: | reinforcement learning feature state sate discretization RBF function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
