%A XU Pingan, TANG Yan, SHI Jiaokai, ZHANG Huirong
%T *K*-Means clustering algorithm based on the Schrödinger equation
%0 Journal Article
%D 2016
%J Journal of Shandong University(Engineering Science)
%R 10.6040/j.issn.1672-3961.0.2015.056
%P 34-41
%V 46
%N 1
%U {http://gxbwk.njournal.sdu.edu.cn/CN/abstract/article_351.shtml}
%8 2016-02-20
%X A new method based on the Schrödinger equation was proposed to find better initial centers. According to the values of potential energy function, initial centers could be selected. Potential energy function value was calculated for each data sample. The data sample with the minimum potential energy function value was placed in the initial cluster centers set. A distance threshold was set to compute distances between the samples from the data set and cluster centers. If a distance was greater than the threshold, this sample would be put into the cluster center and removed from the data set. Otherwise, it would be removed from the data set directly. Repeated this process until the number of initial cluster centers set was equal to *K*. The experimental results showed that the method could find better initial centers and achieve a higher clustering accuracy with fewer iterations. Compared with other methods, the number of iterations could be reduced about 3 times and the accuracy could be increased about 12% by using this method.