Title :
Principal points estimation using mixture distributions
Author :
Ueki, D. ; Matsuura, Saeko ; Suzuki, Hajime
Author_Institution :
Grad. Sch. of Sci. & Technol., Keio Univ., Yokohama, Japan
Abstract :
The k-principal points of a distribution are the k points that optimally partition the distribution. In this paper, we propose a method to estimate principal points from data by using mixture distributions when we have no prior knowledge of the distribution of data. Several simulation results are presented to compare the proposed method with the nonparametric k-means.
Keywords :
data analysis; estimation theory; statistical distributions; data distributions; k-principal points; mixture distributions; nonparametric k-means method; principal points estimation; Clustering algorithms; Data models; Educational institutions; Estimation; Gaussian distribution; Partitioning algorithms; Simulation;
Conference_Titel :
Advanced Computer Science and Information Systems (ICACSIS), 2012 International Conference on
Conference_Location :
Depok
Print_ISBN :
978-1-4673-3026-8
