An application of feature selection for the fuzzy rule based classifier design based on an enlarged Hedge Algebras for high-dimensional datasets

Abstract

The fuzzy rule based classification system (FRBCS) design methods, whose fuzzy rules are in the form of if-then sentences, have been being studied intensively during last years. One of the eminent FRBCS design methods utilizing an enlarged hedge algebras as a formal mechanism to design optimal linguistic terms integrated with their trapezoidal fuzzy sets has been proposed in [12]. As the other methods, a difficult problem needed to be solved is dealing with the high-dimensional and multi-instance datasets. This paper presents an approach to tackle the high-dimensional dataset problem for the FRBCS design method based on an enlarged hedge algebras proposed in [12] by utilizing the feature selection algorithm proposed in [19]. The experimental results over 8 high-dimensional datasets have shown that the proposed method allows saving much execution time than the original one, but retains the equivalent classification performance as well as the equivalent FRBCS complexity.