作者: S. J. Wan , S. K. M. Wong , P. Prusinkiewicz
DOI: 10.1145/45054.45056
关键词: CURE data clustering algorithm 、 Algorithm 、 Data stream clustering 、 Cluster analysis 、 Mathematics 、 Canopy clustering algorithm 、 Ramer–Douglas–Peucker algorithm 、 Fuzzy clustering 、 Correlation clustering 、 Linde–Buzo–Gray algorithm
摘要: A new divisive algorithm for multidimensional data clustering is suggested. Based on the minimization of sum-of-squared-errors, proposed method produces much smaller quantization errors than median-cut and mean-split algorithms. It also observed that solutions obtained from our are close to local optimal ones derived by k-means iterative procedure.