A novel spectral clustering algorithm

In the spectral clustering algorithm, determination of cluster number is a difficult problem. Here an autonomous spectral clustering algorithm is proposed. 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. Finally our algorithm compares with c-means, Ng et.al's algorithm and Francesco et.al's algorithm in the UCI data sets(IRIS data and Wisconsin database). The experiments show our algorithm may get better results.