Feature reduction using fuzzy C-means clustering and Firefly algorithm

Feature selection refers to the elimination of many less informative features. In the proposed method, the Firefly Metaheuristic Algorithm (FMA) selects the features through other dataset features at each stage. Features data are clustered using C-means fuzzy clustering to determine clustering accuracy amounts such as Root-Mean-Square Error (RMSE) to specify how useful these features are and how much these selected features have been able to make classify correctly using clustering based on the dataset as well. Regarding this, the target class is predicted according to the selected features, where the results show the optimal performance of the proposed method. Because of using the combination of FMA and FCM clustering, the optimal centers of each cluster are found quickly, the selected feature sets known as the target class representative have the least error value, and the relationship between features are considered as well by completing the iteration of the algorithm.