ABSTRACT

Spectral clustering and Leader’s algorithm have both been used to identify clusters that are nonlinearly separable in input space. Despite significant research, these methods have remained only loosely related. Sigmoid kernel and polynomial kernel were quite popular for support vector machines due to its origin from clustering. In this paper we are submitting the comparison of above kernel methods after reducing the dimensions using feature functions. For this we have given hand writing data -sets to create and compare the clusters.