A review on advanced c-means clustering models based on fuzzy logic

Clustering represents the task of dividing a set of objects into multiple groups or clusters in such a way that objects in the same group are highly similar to each other, while object in different groups significantly differ from each other. Generally speaking, there are two categories of clustering method: hard clustering refers to the methods that provide a crisp partition in which each object belongs to a single cluster, while soft clustering may assign objects to multiple clusters to a certain extent that can be described by fuzzy membership functions. The so-called c-means clustering models extract a predefined number of cluster prototypes via minimizing a quadratic objective function. This paper proposes to summarize the evolution of c-means clustering methods based on fuzzy logic, from the ISODATA algorithm proposed almost fifty years to the very recent robust clustering algorithms.