Improved FCM algorithm based on the initial clustering center selection

Abstract

Fuzzy C-average algorithm (also known as the FCM algorithm) is a clustering algorithm based on partition. The idea of the algorithm is making is divided into clusters with the similarity between objects as large as possible, and the similarity between objects of different clusters as small as possible. As the algorithm is simple, easy to implement and computer clustering effect, etc., makes this algorithm in many fields has been widely applied. This paper focuses principles of FCM algorithm, the algorithm steps and its problems were described in detail and propose an improved algorithm.