Optimal Control with Reinforcement Learning using Reservoir Computing and Gaussian Mixture

被引：0

作者：

Engedy, Istvan ^{[1
]}

Horvath, Gabor ^{[1
]}

机构：

[1] Budapest Univ Technol & Econ, Dept Measurement & Informat Syst, Budapest, Hungary

来源：

2012 IEEE INTERNATIONAL INSTRUMENTATION AND MEASUREMENT TECHNOLOGY CONFERENCE (I2MTC) | 2012年

关键词：

optimal control; reinforcement learning; ESN; IGMN;

D O I：

暂无

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

Optimal control problems could be solved with reinforcement learning. However it is challenging to use it with continuous state and action spaces, not to speak about partially observable environments. In this paper we propose a reinforcement learning system for partially observable environments with continuous state and action spaces. The method utilizes novel machine learning methods, the Echo State Network, and the Incremental Gaussian Mixture Network.

引用

页码：1062 / 1066

页数：5

共 12 条

[1] Agostini A., 2010, P INT JOINT C NEUR N, P3485
[2] [Anonymous], 2002, approach
[3] [Anonymous], 2005, P 22 INT C MACH LEAR, DOI DOI 10.1145/1102351.1102377
[4] [Anonymous], THESIS U FEDERAL RIO
[5] Berchtold S, 1996, VLDB 1996, P28
[6] Bush K. A., 2008, THESIS COLORADO STAT
[7] Gaussian process dynamic programming
Deisenroth, Marc Peter
Rasmussen, Carl Edward
Peters, Jan
[J]. NEUROCOMPUTING, 2009, 72 (7-9) : 1508 - 1524
[8] Heinen M.R., ARTIFICIAL NEURAL NE, P170
[9] Kernel-based reinforcement learning
Ormoneit, D
Sen, S
[J]. MACHINE LEARNING, 2002, 49 (2-3) : 161 - 178
[10] Sutton R.S., 2017, Introduction to reinforcement learning

← 1 2 →