سهیلا نجفی گوجانی - Department of Electrical Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran
سیدحمید محمودیان - Department of Electrical Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran, Digital Processing and Machine Vision Research Center, Najafabad Branch, Islamic Azad University, Najafabad, Iran

In this paper, a fuzzy rule extraction system is designed and the idea of ​​using this method in designing a fuzzy rule database is considered. First, the fuzzy backup vector machine is designed based on the clustering method. In this case, the backup vector machine will perform better in noise data or out-of-range data, and as a result, the rules extracted from it seem to be more accurate. The fuzzy backup vector machine model is simulated and then its backup vectors are used to derive fuzzy rules. The extracted rules are fuzzy rules that are obtained in the form of "if-then". The rules are extracted using two models of simple support vector machine and fuzzy support vector machine, which will show that the rules extracted from fuzzy backup vector machine in most cases have better accuracy in classification than ordinary backup vector machine. The number of extracted rules is an important parameter in the accuracy of the operation of this database, the number of rules extracted in the fuzzy method is less in most cases.

Support vector machine, fuzzy rule base, fuzzy clustering, fuzzy classification., ماشین‌بردارپشتیبان, پایگاه‌قواعدفازی, بردارپشتیبان, خوشه‌بندی‌فازی, طبقه‌بندهای‌فازی