基于Rough Set的任务复杂度与协同模型研究

Research on Task-related Complexity and Cooperative ModelBased on Rough Sets

基于RoughSet(RS)理论 ,在知识的层次上 ,给出了相关任务的知识表达以及任务复杂度的概念。针对多Agent系统 ,提出了衡量Agent基于知识水平的能力、协同的冗余度等方法。运用粗糙度、信息熵和知识这三个指标 ,讨论了协同的可能条件以及必然条件。对于协同满足可能条件而不满足必然条件的情况下 ,提出了进行学习的最小知识集的概念。在此基础上 ,给出了相关任务T的静态协同模型与算法。最后 ,指出了有待解决的问题。

Task related knowledge representation and complexity concepts are discussed based on Rough Sets theory ,and methods of measuring ability and redundant of Agents are presented on knowledge level in a multi agent system. By those concepts of rough measurement, information entropy and knowledge, cooperative necessary conditions and sufficient conditions are provided. A task related static cooperative model and some problems to be solved are discussed in the end.