A Modified Spectral Clustering Algorithm Based on NJW

Spectral clustering has become one of the most popular modern clustering algorithms because it has "global" optimal solution compared with traditional clustering methods. In this paper, we propose a modified NJW algorithm which is based on matrix perturbation theory and can be easily implemented. In addition, the algorithm can estimate the parameter k and in turn achieve appropriate clusters. The experimental results on a number of challenging clustering problems show that it has good performance.