NCestimation_V2

  The NCT includes 4 External validity indices and 8 internal validity indices, and the sub-routine validity_Index.m is designed to use them. This tool is suitable for the research work such as the performance comparison of different indices on estimation of the number of clusters, algorithm design by improving part codes of this toolbox. i) External validity indices when true class labels are known: Rand index Adjusted Rand index Mirkin index Hubert index ii) Internal validity indices when true class labels are unknown: Silhouette Davies-Bouldin Calinski-Harabasz Krzanowski-Lai Hartigan weighted inter- to intra-cluster ratio Homogeneity Separation iii) Others Error rate (compared with true labels) System Evolution: it is used to estimate the number of clusters and give separable degrees between clusters.