A clustering method by code vectors considering attractive and repulsive force based on maximum distance of k neighbors
Hiroki Imamura, Makoto Fujimura, Hideo Kuroda · 2008
When noise data is included, the clustering method based on code vectors considering attractive and repulsive force can not precisely classify data. In this paper, we propose the clustering method based on code vectors considering attractive and repulsive force which can precisely classify data even when noise data is included.