Autonomous spectral clustering

Cluster number is chosen in advance in the traditional spectral clustering algorithms. Here a spectral clustering algorithm is proposed to decide autonomously the cluster number, which eliminates the drawbacks of two kinds of spectral clustering algorithm methods. Eigengap is used to discover the clustering stability and decide automatically the cluster number, which is proved theoretically the rationality of cluster number. A kernel based fuzzy c-means is introduced to spectral clustering algorithm to separate clusters. Finally the experiments show our algorithm may get better results than c-means, Ng et al.'s algorithm and Francesco et al.'s algorithm in the UCI data sets.