Classification with Fuzzy OWA Distance

Abstract

OWA (Ordered Weighted Averaging) Distance Based CxK Nearest Neighbor Algorithm (CxK-NN) via L-R fuzzy data is performed with two different fuzzy metric measures. We use fuzzy metric defined by Diamond and a weighted dissimilarity measure composed by spread distances and center distances in order to evaluate the effects of different metric measures. K neighbors are considered for each class and the algorithm perform OWA operator in order to calculate the distance between being classified fuzzy point and its K-nearest set. It is observed that the OWA distance behavior by changing its weights as inter-cluster distance approaches single, complete, and average linkages. The performance of this novel approach is evaluated by using n-fold cross validation. After experiments with well-known three classification dataset, it is observed that single linkage approach by using two different metric measures has significant different results.