期刊文献+

多代理强化学习在智能教学系统中的应用

Applications of Multi-agent and Reinforcement Learning in ITS
下载PDF
导出
摘要 教学的个性化和智能化是智能教学系统研究的重点和难点。文章采用智能代理技术模拟系统中学生的智能和行为方式,将强化学习理论应用于多代理体,设计了结合资格迹理论的强化学习算法,并用以生成和调整适合于每个学生个体的教学内容和教学策略。多代理体技术实现了教学的个性化,强化学习算法使得教学策略具有智能化。实验结果表明,新的算法较原有算法更为有效。 Personalized and intelligent are two focuses on the Intelligent Tutoring system(ITS).The intelligence and behavior of students are simulated by intelligent agent technology in this system,the reinforcement learning theory is applied to multi-agent body,and a new reinforcement learning algorithm is designed combining with the theory of eligibility trace.The new algorithm is used to build and develop teaching content and teaching strategies that appropriate for each student.Multi-Agent technology achieves the personalized in ITS,and reinforcement learning algorithm makes teaching strategies with the intelligent.The experimental results show that the new algorithm is better than the original algorithm.
作者 李洋
机构地区 合肥工业大学
出处 《计算机与数字工程》 2010年第5期78-80,174,共4页 Computer & Digital Engineering
关键词 多代理体 强化学习 智能教学系统 multi-agent reinforcement learning intelligent tutoring system
  • 相关文献

参考文献8

二级参考文献17

  • 1田盛丰,人工智能与知识工程,1999年
  • 2M Wooldridge, et al. Intelligent Agents: Theory and Practice [J]. Knowledge Engineering Review, 1995. 10(2): 115-152.
  • 3(美)约翰·H·霍兰著 周晓牧 韩晖译.隐秩序--适应性造就复杂性[M].上海:上海科技教育出版社,2000..
  • 4Sutton R S, Barto A G. Reinforcement learning: An introduction[M]. London: Cambridge MIT Press, 1998.
  • 5Watkins C J. Learning with delayed rewards[D]. London: Cambridge University Psychology Depart-ment, 1989.
  • 6Singh S, Sutton R S. Reinforcement learning with replacing eligibility traces[J]. Machine Learning, 1996, 22:123-158.
  • 7Cichosz P. Truncating temporal differences: On the efficient implementation of TD(λ) for reinforcement learning[J]. Journal on Artificial Intelligence,1995(2):287-318.
  • 8Singh S, Cohn D. How to dynamically merge Markov decision processes[J]. Advances in Neural Informa-tion Processing Systems, 1998, 10:118-131.
  • 9Singh S, Jaakkola T, Littman M L, et al. Convergence results for single-step on-policy reinfor-cement-learning algorithms[J]. Machine Learning, 2000,26:264-281.
  • 10Sutton R S.Learning to predict by the methods of temporal differences[J]. Machine Learning,1988(3):9-44.

共引文献49

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