AN INVESTIGATION OF FUZZY MODEL USING AIC

In this paper we suggest the ''goodness'' of models using the informat ion criterion AIC. The information criterion AIC is a statistic to est imate the badness of models. When we usually make the fuzzy rules, we aim to minimize inference error and the number of rules. But these con ditions are the criteria to acquire an optimum rule-model by using the training data. In the general case of fuzzy reasoning, we aim to mini mize the inference error fur not only given training data, but also un known data. So we have introduced a new information criterion based on AIC into the appraised criterion fur estimating the acquired fuzzy ru les. Experimental results are given to show the validity of using AIC.