Learning Fuzzy Cognitive Maps by a Hybrid Method UsingNonlinear Hebbian Learning and Extended Great DelugeAlgorithm

23rd Midwest Artificial Intelligence and Cognitive Science Conference 2012

Zhaowei Ren

Others  :  http://ceur-ws.org/Vol-841/submission_27.pdf

来源: CEUR

【 摘 要 】

Fuzzy Cognitive Maps (FCM) is a technique to representmodels of causal inference networks. Data driven FCMlearning approach is a good way to model FCM. Wepresent a hybrid FCM learning method that combinesNonlinear Hebbian Learning (NHL) and Extended GreatDeluge Algorithm (EGDA), which has the efficiency ofNHL and global optimization ability of EGDA. Wepropose using NHL to train FCM at first, in order to getclose to optimization, and then using EGDA to make modelmore accurate. We propose an experiment to test the accuracy and running time of our methods.