摘要
Agent具有自私的特性.为了在多agent系统学习时使学习效果体现最大的群体利益,提出了一种改进的多agent学习方法.该学习方法改进了基于基组织共识学习的学习方法,在多agent学习时考虑了主体的可信度问题,从而使最后的学习结果更具真实性.
Agent is selfish. In order to get maximal system benefit through MAS learning, an improved approach of learning in Multi - agent system is proposed. In this approach, the learning method which based on organization - structure by considering the credit of the agent, and gets a better outcome.
出处
《辽宁大学学报(自然科学版)》
CAS
2007年第1期81-84,共4页
Journal of Liaoning University:Natural Sciences Edition
基金
江苏省教育厅科研项目(02KJD520004)
关键词
多AGENT
组织
学习
multi - agent
organization
learning