EM clustering analysis of diabetes patients basic diagnosis index.
Cai Wu, Jeffrey R. Steinbauer, Grace m. Kuo · PubMed · 2005
Cluster analysis can group similar instances into same group and different instances into different groups. It assigns classes to samples without known the classes in advance. EM clustering algorithm can find number of distributions of generating data and build "mixture models". It identifies groups that are either overlapping or varying sizes and shapes. In this project, by using EM in Weka system, diabetes patient basic diagnosis index data have been analyzed for clustering.