Cluster validity measurement techniques

Abstract:- Clustering is a process of discovering groups of objects such that the objects of the same group are similar, and the objects belonging to different groups are dissimilar. Several research fields deal with the problem of clustering: for example pattern recognition, data mining, machine learning. A number of algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input parameters. Therefore it is very important to evaluate the result of the clustering algorithms. It is difficult to define whether a clustering result is acceptable or not, thus several clustering validity techniques and indices have been developed. This paper deals with the problem of clustering validity. The most commonly used validity indices are introduced and explained, and they are compared based on experimental results