Interval Type-2 fuzzy C-Means approach to collaborative clustering

There have been numerous studies on using the FCM algorithm in clustering and collaboration clustering, especially in data analysis, data mining and pattern recognition. In this study, we present new methods involving interval Type-2 fuzzy sets to realize collaborative clustering. Data in which the clustering results realized at one data site impact clustering carried out at other data sites. Those methods endowed with interval type-2 fuzzy sets help cope with uncertainties present in data. The experiment with weather data sets has shown better results in comparison with the previous approaches.

I. Introduction

Clustering is used to detect a sound structure or patterns in the data set, in which objects positioned within the cluster level data show a substantial level of similarity. This unsupervised technique has a long history in machine learning, pattern recognition, data mining, and many algorithms have been exploited in various applications. Clustering algorithms comes in numerous varieties including k - means and its variants [1]–[2], and a family of Fuzzy C-Mean (FCM) [3]–[5].