Publisher review:SysEvolution - output optimal number of clusters, border distances and separable degrees between clusters System Evolution (SE) method is developed to estimate the number of clusters (NC) based on PAM (partitioning around medoids), a robust clustering algorithm or a robust ¡°version¡± of K-means, and it provides the intuitive information about the distances and separable degree between clusters. It is a useful tool to discover and understand the cluster structure of a data set.The SE pays attention to the distinguishable features of obvious distance between the borders of clusters or obvious sparse data points in overlapping region of clusters. Hence, it is applicable to the data with well-separated clusters and slightly overlapping clusters.The SE is developed for the PAM clustering algorithm. Your using and tests on more algorithms as well as comments are helpful and welcome. Requirements: ยท MATLAB Release: R13
SysEvolution 3.1 is a Matlab script for Statistics and Probability scripts design by Kaijun Wang.
It runs on following operating system: Windows / Linux / Mac OS / BSD / Solaris.
SysEvolution - output optimal number of clusters, border distances and separable degrees between clusters
Operating system:Windows / Linux / Mac OS / BSD / Solaris