摘要
模糊认知图(Fuzzy Cognitive Map,FCM)是知识表示和推理的有效工具。本论文主要研究一种FCM的系统分析方法:对于复杂FCM,首先将其分割成基本FCM,然后对每一个基本FCM建立关键节点集合,通过确定关键节点的状态,来确定其它节点的状态,再通过基本FCM之间的连接关系确定全部节点的状态。本文提出了基本FCM,关键节点等概念,并在此基础上进行了相关结果的证明,给出了通用算法。最后通过实例来说明算法的有效性。
Fuzzy Cognitive Map (FCM) is a powerful tool for representing structured human knowledge and causal inference. This paper proposes an approach for analyzing FCM systematically. In this method, a complicated FCM is divided into several basic FCMs. Then, the key vertices set is constructed in every basic FCM. The state of every vertex in the basic FCM is determined by the states of key vertices. Several theorems are proved to guarantee the mothod and an algorithm is proposed for implementation. The algorithm is illustrated by an example.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1999年第2期143-151,共9页
Pattern Recognition and Artificial Intelligence
基金
The Young Asia Scholar Award of Melborne University
国家自然科学基金
航空科学基金
