期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

基于预测状态表示模型和稀疏分布记忆的多观测系统预测

Prediction of multi-observation system based on predictive state representation and sparse distribute memory
下载PDF
导出
摘要 提出了一种新型的PSR建模方法,该方法建立针对复杂多观测系统的近似预测模型S-PSR,将系统中的检验和经历依据归属关系进行归类划分,利用稀疏分布记忆(SDM)存储结构进行模型当前状态保存和状态更新,实现了对多观测系统复杂数据的处理。实验表明,该近似模型相比其他模型具有更好的预测准确性。 this paper proposed a new model creation method for PSR. The method constructed the approximate prediction model for complex multi-observations system-S-PSR. The new model divided the tests and the histories for classification based on relation. Furthermore the model finished the state saving and updating by using SDM and realized the process of complex data on multi-observation system. The result shows that the proposal model is better than other model in prediction error.
作者 汪庆淼 鞠时光
机构地区 江苏大学计算机科学与通信工程学院 苏州大学计算机科学与技术学院
出处 《计算机应用研究》 CSCD 北大核心 2012年第8期2988-2990,共3页 Application Research of Computers
基金 国家自然科学基金资助项目(60773049)
关键词 多观测系统 预测状态表示 稀疏分布记忆 系统模型 multi-observation system predictive state representation （PSR） sparse distribution memory （ SDM ） system modeling
分类号 TP29 [自动化与计算机技术—检测技术与自动化装置]
  • 相关文献

参考文献11

  • 1SINGH S, JAMES M R, RUDARY M R. Predictive state representa- tions: a new theory for modeling dynamical systems [ C ]//Proc of the 20th Conference on Uncertainty in Artificial Intelligence. Banff: AUAI Press,2004:512-519.
  • 2BOWLING M, McCRACKEN P, JAMES M,et al. Learning predictive state representations using non-blind policies [ C ]//Proc of the 23 rd International Conference on Machine Learning. New York: ACM, 2006 : 129-136.
  • 3TALVITIE E, SINGH S. Simple local models for complex dynamical Systems [ C ]//Proc of the 22nd Annual Conference on Neural Informa- tion Processing Systems. Vancouver :AUAI Press,2008:8-11.
  • 4DINCULESCU M, PRECUP D. Approximate predictive representa- tions of partially observable systems [ C ]//Proe of the 27th Internation- al Conference on Machine Learning. New York :ACM,2010:859-902.
  • 5WOLFE B, JAMES M R, SINGH S. Learning predictive state repre- sentations in dynamical systems without reset [ C ]//Proc of the 22nd International Conference on Machine Learning. New York: ACM, 2005:980-987.
  • 6MENG Hong-ying, APIAH K, HUNTER A, et al. A modified sparse distributed'memory model for extracting clean patterns from noiay in- puts [ C ]//Proc of IEEE International Joint Conference on Neural Net- works. Atlanta: IEEE Press,2009:2084- 2089.
  • 7WOLFE B, JAMES M R, SINGH S. Approximate predictive state representations [ C ]//Proc of the 7th International Joint Conference on Autonomous Agents and Multiagent Systems. Estoril: AAAI Press, 2008:363-370.
  • 8BOOTS B, SIDDIQI S M, GORDON G J. Closing the learning-plan- ning loop with predictive state representations [ J ]. The International Journal Robotics ,2011,30(7 ) :954-966.
  • 9WINGATE D, SINGH S. On discovery and learning of models with predictive state representations of state for agents with continuous ac- tions and observations [ C ]//Proc of the 6th International Joint Confer- ence on Autonomous Agents and Muhiagent Systems. New York: ACM ,2007 : 187.
  • 10TALVITIE E, SINGH S. Maintaining predictions over time without a model[ C ]//Proc of the 21 st International Joint Conference on Artifi- cial Intelligence. Pasadena: AAAI Press ,2009 : 1249-1254.
计算机应用研究
2012年 第8期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部