Many real-world and man-made objects are symmetry. Therefore, it is reasonable to assume that some kinds of symmetry may exist in data clusters. The most common type of symmetry is line symmetry. In this paper, we propose a line symmetry distance measure. Based on the proposed line symmetry distance, a modified version of the K-means algorithm can be used to partition data into clusters with different geometrical shapes. Several data sets are used to test the performance of the proposed modified version of the K-means algorithm incorporated with the line symmetry distance.
International Journal of Innovative Computing, Information and Control 7(8), pp.5027-5043