A mixture model-based on-line CEM algorithm

Authors:
Allou Samé;Gérard Govaert;Christophe Ambroise
Affiliations:
Département Génie Informatique, HEUDIASYC, UMR CNRS 6599, Université de Technologie de Compiègne, Compiègne Cedex;Département Génie Informatique, HEUDIASYC, UMR CNRS 6599, Université de Technologie de Compiègne, Compiègne Cedex;Département Génie Informatique, HEUDIASYC, UMR CNRS 6599, Université de Technologie de Compiègne, Compiègne Cedex
Venue:
IDA'05 Proceedings of the 6th international conference on Advances in Intelligent Data Analysis
Year:
2005

Citing 2
Cited 1

A Classification EM algorithm for clustering and two stochastic versions

Computational Statistics & Data Analysis - Special issue on optimization techniques in statistics
On-line learning and stochastic approximations

On-line learning in neural networks

On online high-dimensional spherical data clustering and feature selection

Engineering Applications of Artificial Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

An original on-line mixture model-based clustering algorithm is presented in this paper. The proposed algorithm is a stochastic gradient ascent derived from the Classification EM (CEM) algorithm. It generalizes the on-line k-means algorithm. Using synthetic data sets, the proposed algorithm is compared to CEM and another on-line clustering algorithm. The results show that the proposed method provides a fast and accurate estimation when mixture components are relatively well separated.