作者: D. Simovici , N. Singla , M. Kuperberg
关键词: Cluster analysis 、 Algorithm 、 FLAME clustering 、 Single-linkage clustering 、 Pattern recognition 、 Mathematics 、 Data stream clustering 、 CURE data clustering algorithm 、 Correlation clustering 、 Fuzzy clustering 、 k-medians clustering 、 Artificial intelligence
摘要: We present an algorithm/or clustering nominal data that is based on a metric the set of partitions finite objects; this defined starting from lower valuation lattice partitions. The proposed algorithm seeks to determine partition such total distance between and determined by attributes objects has local minimum. resulting quite stable relative ordering objects.