An Adaptive Incremental Clustering Method based on the Growing Neural Gas Algorithm

Mohamed-Rafik Bouguelia, Yolande Belaïd, Abdel Belaïd

2013

Abstract

Usually, incremental algorithms for data streams clustering not only suffer from sensitive initialization parameters, but also incorrectly represent large classes by many cluster representatives, which leads to decrease the computational efficiency over time. We propose in this paper an incremental clustering algorithm based on ”growing neural gas” (GNG), which addresses this issue by using a parameter-free adaptive threshold to produce representatives and a distance-based probabilistic criterion to eventually condense them. Experiments show that the proposed algorithm is competitive with existing algorithms of the same family, while maintaining fewer representatives and being independent of sensitive parameters.

Download


Paper Citation


in Harvard Style

Bouguelia M., Belaïd Y. and Belaïd A. (2013). An Adaptive Incremental Clustering Method based on the Growing Neural Gas Algorithm . In Proceedings of the 2nd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-8565-41-9, pages 42-49. DOI: 10.5220/0004256600420049

in Bibtex Style

@conference{icpram13,
author={Mohamed-Rafik Bouguelia and Yolande Belaïd and Abdel Belaïd},
title={An Adaptive Incremental Clustering Method based on the Growing Neural Gas Algorithm},
booktitle={Proceedings of the 2nd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2013},
pages={42-49},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004256600420049},
isbn={978-989-8565-41-9},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 2nd International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - An Adaptive Incremental Clustering Method based on the Growing Neural Gas Algorithm
SN - 978-989-8565-41-9
AU - Bouguelia M.
AU - Belaïd Y.
AU - Belaïd A.
PY - 2013
SP - 42
EP - 49
DO - 10.5220/0004256600420049