Initialization of Iterative Refinement Clustering Algorithms Export

@proceedings{citeulike:3562, abstract = {Iterative refinement clustering algorithms (e.g. K-Means, EM) converge to one of numerous local minima. It is known that they are especially sensitive to initial conditions. We present a procedure for computing a refined starting condition from a given initial one that is based on an efficient technique for estimating the modes of a distribution. The refined initial starting condition leads to convergence to \"better\" local minima. The procedure is applicable to a wide class of clustering...}, author = {Fayyad, Usama M. and Reina, Cory and Bradley, Paul S.}, booktitle = {Knowledge Discovery and Data Mining}, citeulike-article-id = {3562}, citeulike-linkout-0 = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.54.3469}, keywords = {commented, image-processing, machine-learning, sampling}, pages = {194--198}, posted-at = {2007-11-05 04:57:30}, priority = {2}, title = {Initialization of Iterative Refinement Clustering Algorithms}, url = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.54.3469}, year = {1998} }

TY - CONF ID - citeulike:3562 L3 - citeulike-article-id:3562 N2 - Iterative refinement clustering algorithms (e.g. K-Means, EM) converge to one of numerous local minima. It is known that they are especially sensitive to initial conditions. We present a procedure for computing a refined starting condition from a given initial one that is based on an efficient technique for estimating the modes of a distribution. The refined initial starting condition leads to convergence to "better" local minima. The procedure is applicable to a wide class of clustering... EP - 198 TI - Initialization of Iterative Refinement Clustering Algorithms T2 - Knowledge Discovery and Data Mining SP - 194 KW - commented KW - image-processing KW - machine-learning KW - sampling AU - Fayyad, Usama AU - Reina, Cory AU - Bradley, Paul PY - 1998/// UR - http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.54.3469 ER -