A hybrid lter-based feature selection method via hesitant fuzzy and rough sets concepts

High dimensional microarray datasets are dicult to classify since they have many features with small number of instances and imbalanced distribution of classes. This paper proposes a lter-based feature selection method to improve the classication performance of microarray datasets by selecting the signicant features. Combining the concepts of rough sets, weighted rough set, fuzzy rough set and hesitant fuzzy sets for developing an effective algorithm is the main contribution of this paper. The mentioned method has two steps, in the rst step, four discretization approaches are applied to discretize continuous datasets. Also, a primary subset of features is selected by combining of weighted rough set dependency degree and information gain via hesitant fuzzy aggregation approach. In the second step, a signicance measure of features (dened by fuzzy rough concepts) is employed to remove redundant features from primary set. The Wilcoxon Signed Ranked test (A Non-parametric statistical test) is conducted for comparing the presented method with ten feature selection methods across seven datasets. The results of experiments show that the proposed method is able to select a signicant subset of features and it is an eective method in the literature in terms of classication performance and simplicity.