ASSESSMENT OF THE EFFECT OF NOISE ON AN UNSUPERVISED FEATURE SELECTION METHOD FOR GENERATIVE TOPOGRAPHIC MAPPING

Alfredo Vellido, Jorge S. Velazco

2008

Abstract

Unsupervised feature relevance determination and feature selection for dimensionality reduction are important issues in many clustering problems. An unsupervised feature selection method for general Finite Mixture Models was recently proposed and subsequently extended to Generative Topographic Mapping (GTM), a nonlinear manifold learning constrained mixture model for data clustering and visualization. Some of the results of a previous preliminary assessment of this method for GTM suggested that its performance may be affected by the presence of uninformative noise in the dataset. In this brief study, we test in some detail such limitation of the method.

References

  1. Bishop, C., Svensén, M., and Williams, C. (1998). Developments of the generative topographic mapping. In Neurocomputing. 21(1-3), pp. 203-224.
  2. Bishop, C., Svensén, M., and Williams, C. (1999). Gtm: The generative topographic mapping. In Neural Computation. 10(1), pp. 215-234.
  3. Law, M., Figueiredo, M., and Jain, A. (2004). Simultaneous feature selection and clustering using mixture models. In IEEE T. Pattern Anal. 26(9), pp. 1154-1166.
  4. McLachlan, G. and Peel, D. (1998). Finite mixture models. John Wiley-Sons, New York.
  5. Olier, I. and Vellido, A. (2006). Time series relevance determination through a topology-constrained hidden markov model. In Proc. of the 7th International Conference on Intelligent Data Engineering and Automated Learning (IDEAL 2006). LNCS 4224, 40-47. Burgos, Spain.
  6. Olier, I. and Vellido, A. (2008). On the benefits for model regularization of a variational formulation of gtm. In in Proceedings of the International Joint Conference on Neural Networks (IJCNN 2008). in press.
  7. Vellido, A. (2005). Preliminary theoretical results on a feature relevance determination method for generative topographic mapping. In Technical Report LSI-05-13- R. Universitat Politecnica de Catalunya, Barcelona, Spain.
  8. Vellido, A. (2006). Assessment of an unsupervised feature selection method for generative topographic mapping. In 16th International Conference on Artificial Neural Networks. LNCS 4132, 361-370. Athens, Greece.
  9. Vellido, A., El-Deredy, W., and Lisboa, P. (2003). Selective smoothing of the generative topographic mapping. In IEEE T. Neural Network. 14(4), pp. 847-852.
  10. Vellido, A., Lisboa, P., and Vicente, D. (2006). Robust analysis of mrs brain tumour data using t-gtm. In Neurocomputing. 69(7-9), pp. 754-768, 2006.
  11. 3 new features - Std. dev = 0.2
Download


Paper Citation


in Harvard Style

Vellido A. and S. Velazco J. (2008). ASSESSMENT OF THE EFFECT OF NOISE ON AN UNSUPERVISED FEATURE SELECTION METHOD FOR GENERATIVE TOPOGRAPHIC MAPPING . In Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS, ISBN 978-989-8111-37-1, pages 424-430. DOI: 10.5220/0001681204240430


in Bibtex Style

@conference{iceis08,
author={Alfredo Vellido and Jorge S. Velazco},
title={ASSESSMENT OF THE EFFECT OF NOISE ON AN UNSUPERVISED FEATURE SELECTION METHOD FOR GENERATIVE TOPOGRAPHIC MAPPING},
booktitle={Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS,},
year={2008},
pages={424-430},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001681204240430},
isbn={978-989-8111-37-1},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS,
TI - ASSESSMENT OF THE EFFECT OF NOISE ON AN UNSUPERVISED FEATURE SELECTION METHOD FOR GENERATIVE TOPOGRAPHIC MAPPING
SN - 978-989-8111-37-1
AU - Vellido A.
AU - S. Velazco J.
PY - 2008
SP - 424
EP - 430
DO - 10.5220/0001681204240430