The USC Andrew and Erna Viterbi School of Engineering USC Signal and Image Processing Institute USC Ming Hsieh Department of Electrical Engineering University of Southern California

Technical Report USC-SIPI-261

“A Fuzzy Classifier That Uses Both Crisp Samples and Linguistic Knowledge”

by Wen Wei and Jerry M. Mendel

June 1994

In this paper we develop a general structure of a fuzzy logic (FL) classifier that is capable of using both numerical data and linguistic information. By using fuzzy inference, we are able to handle numerical data and linguistic information in a unified framework. We show that the FL classifier includes the Bayes classifier as a special case. Our experimental results show that the FL classifier, when using linguistic information, can perform better than probabilistic classifiers that do not use linguistic information.

To download the report in PDF format click here: USC-SIPI-261.pdf (0.9Mb)