Fast algorithm for finding true number of clusters. applications to control systems
Mikhail Morozkov, Oleg Nikolaevich Granichin, Zeev Volkovich, Xuming Zhang · 2012
One of the most difficult problems in cluster analysis is the identification of the number of groups in a given data set. In this paper we offer the randomized approach in the rate distortion framework. A randomized algorithm has been suggested to allocate this position. The scenario approach is used to significantly reduce the computational complexity. With ability to determine the true number of clusters and perform clustering in real-time operational mode we outline several applications in control systems and decision-making problems that can benefit from algorithm in question essentially. We also provide simulation results to show considerable speed optimization with guaranteed level of probability.