ONLINE LEARNING OF GAUSSIAN MIXTURE MODELS - A Two-Level Approach

Arnaud Declercq, Justus H. Piater

Abstract

We present a method for incrementally learning mixture models that avoids the necessity to keep all data points around. It contains a single user-settable parameter that controls via a novel statistical criterion the trade-off between the number of mixture components and the accuracy of representing the data. A key idea is that each component of the (non-overfitting) mixture is in turn represented by an underlying mixture that represents the data very precisely (without regards to overfitting); this allows the model to be refined without sacrificing accuracy.

Download


Paper Citation


in Harvard Style

Declercq A. and H. Piater J. (2008). ONLINE LEARNING OF GAUSSIAN MIXTURE MODELS - A Two-Level Approach . In Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: OPRMLT, (VISIGRAPP 2008) ISBN 978-989-8111-21-0, pages 605-611. DOI: 10.5220/0001090506050611


in Bibtex Style

@conference{oprmlt08,
author={Arnaud Declercq and Justus H. Piater},
title={ONLINE LEARNING OF GAUSSIAN MIXTURE MODELS - A Two-Level Approach},
booktitle={Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: OPRMLT, (VISIGRAPP 2008)},
year={2008},
pages={605-611},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001090506050611},
isbn={978-989-8111-21-0},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Third International Conference on Computer Vision Theory and Applications - Volume 1: OPRMLT, (VISIGRAPP 2008)
TI - ONLINE LEARNING OF GAUSSIAN MIXTURE MODELS - A Two-Level Approach
SN - 978-989-8111-21-0
AU - Declercq A.
AU - H. Piater J.
PY - 2008
SP - 605
EP - 611
DO - 10.5220/0001090506050611