The Global Interval Type-2 Fuzzy C-Means clustering algorithm

Abstract

The Interval Type-2 Fuzzy C-Means clustering algorithm (IT2FCM) is one of the algorithms for clustering based on optimizing a target function, which is sensitive to initial conditions. Aiming at this problem, we propose the Global Interval Type-2 Fuzzy C-Means (GIT2FCM) clustering algorithm which is not only independent on any initial conditions by dynamically increasing cluster center, but also achieves optimal clustering purposes by global search. Experiments show that the Global Interval type-2 Fuzzy C-Means clustering algorithm has better experiment results by overcoming sensibility to initial value, and improves the accuracy of clustering.