F. Meimandi Parizi - School of Electrical and Computer Engineering Shiraz University Shiraz, Iran
E.G. Mansoori - School of Electrical and Computer Engineering Shiraz University Shiraz, Iran

In this paper we have employed an interpretable FRBCS for classification of protein structural classes, since there is a lack of comprehensible classification systems in this field. In fact the aim of a FRBCS is improvement in the performance of systems and in some studies comprehensibility has also been taken into account. In addition, a fuzzy system which generates a rule base with fewer and shorter general rules provides more comprehensible system. The proposed fuzzy classification method generalizes antecedent fuzzy sets, hence generates interpretable fuzzy rules for biologists to predict protein structural classes. Prediction of protein structural classes is an important problem in protein science, as it provides useful information toward determination of overall protein structure and functions. The resulting classification system achieved high prediction accuracy with jackknife cross-validation test for two benchmark datasets, and the generated rule base is compact and comprehensible

Fuzzy classification system; comprehensibility; predicting protein structural classes