<![CDATA[This systemuses a Gaussian Means clustering for which the method is trying to determine the number of clusters either gradually or in stages to see if a cluster is in a state normally distributed or not. If the data considered not normal, it will be seen whether the cluster is still possible to be split two cluster. The workings of the G-Means method begins with determining the alpha error for testing and to determine the number of clusters to be formed in the initial clustering have two centroids. Anderson darling test to see if the cluster necessary splitting or not. The test results show that the Gaussian Means method has a clustering waters can do well, but the green areas and the woke up area not. Weaknesses contained in this application is the image that will adversely affect the clustering results]]>
Copyrights © 2013
