Intuitionistic Fuzzy c-Ordered Means Clustering Algorithm

Abstract

Atanassov intuitionistic fuzzy set (AIFS) has the capability to deal with various uncertain situations, so its popularity among researchers is quite high. It has been observed that Euclidean distance measure based AIFS clustering algorithms perform well on imprecise datasets. The performance of Euclidean measure based clustering algorithms deteriorates due to the presence of outliers/noise within a dataset. In the paper, an extension of the algorithm given by Leski, Jacek M. [Fuzzy Sets and Systems, 286 (2016): 114-133] is proposed as intuitionistic fuzzy <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$c$ </tex-math></inline-formula> -ordered means clustering algorithm. This paper analyses the functionality of clustering algorithms over outliers/noises based datasets. In IFCOM, an alternate of Euclidean distance known as Loss function is used. Moreover, IFCOM uses intuitionistic fuzzy OWA to combat the ill effects of the noises and outliers. The proposed algorithm exploits a typicality function based weighing ordering approach. The approach assigins lower weights to the outliers. Hence, the catastrophic behavior is not observed in IFCOM while dealing outliers possessing synthetic and UCI machine learning datasets.