| 几种agent强化学习方法的比较研究 |
| |
| 引用本文： | 吴元斌.几种agent强化学习方法的比较研究[J].数字社区&智能家居,2008(5):774-776. |
| |
| 作者姓名： | 吴元斌 |
| |
| 作者单位： | 重庆三峡学院数学与计算机科学学院,重庆404000 |
| |
| 摘 要： | 强化学习使agent具有在线自主学习能力，该文介绍了MDP模型下的自适应动态规划、时序差分学习、Q-学习等几种典型agent强化学习方法，并从基本思想、学习内容、收敛速度、可扩展性等方面对它们进行了对比分析：
|
| 关 键 词： | MDP 自适应动态规划 时序差分学习 Q-学习 |
Comparative Analysis of Some Agent Reinforcement Learning Methods |
| |
| WU Yuan-bin.Comparative Analysis of Some Agent Reinforcement Learning Methods[J].Digital Community & Smart Home,2008(5):774-776. |
| |
| Authors: | WU Yuan-bin |
| |
| Affiliation: | WU Yuan-bin (College of Mathematics and Computer Science, Chongqing Three Gorges University, Chongqing 404000, China) |
| |
| Abstract: | Reinforcement learning enables agent online learning autonomously, some typical reinforcement methods such as ADP, TD learning, Q-learning which based on MDP model are introduced in this paper. And comparative analysis from the aspects such as the basic idea, learning content, convergence rate, and extendibility has carried out. |
| |
| Keywords: | MDP Adaptive Dynamic Programming TD Learning Q-Learning |
| 本文献已被 维普 等数据库收录！ |
