Sample selection via clustering to construct support vector-like classifiers
Authors:
- M. MartinezMMM. Martinez
- I. MoraI. Mora
- M. VaquezMVM. Vaquez
- J.-L. SanchoJSJ.-L. Sancho
- A.R. Figueiras-VidalAFA.R. Figueiras-Vidal
- 1 November 1999
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks
- Vol. 10 (6) , 1474-1481
Abstract
Explores the possibility of constructing RBF classifiers which, somewhat like support vector machines, use a reduced number of samples as centroids, by means of selecting samples in a direct way. Because sample selection is viewed as a hard computational problem, this selection is done after a previous vector quantization: this way obtains also other similar machines using centroids selected from those that are learned in a supervised manner. Several forms of designing these machines are considered, in particular with respect to sample selection; as well as some different criteria to train them. Simulation results for well-known classification problems show very good performance of the corresponding designs, improving that of support vector machines and reducing substantially their number of units. This shows that our interest in selecting samples (or centroids) in an efficient manner is justified. Many new research avenues appear from these experiments and discussions, as suggested in our conclusions.Keywords
This publication has 19 references indexed in Scilit:
- Cost functions to estimate a posteriori probabilities in multiclass problemsIEEE Transactions on Neural Networks, 1999
- The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the networkIEEE Transactions on Information Theory, 1998
- Comparing support vector machines with Gaussian kernels to radial basis function classifiersIEEE Transactions on Signal Processing, 1997
- Support-vector networksMachine Learning, 1995
- The Nature of Statistical Learning Theory1995
- Pedagogical pattern selection strategiesNeural Networks, 1994
- A training algorithm for optimal margin classifiers1992
- International application of a new probability algorithm for the diagnosis of coronary artery diseaseThe American Journal of Cardiology, 1989
- Simplifying decision treesInternational Journal of Man-Machine Studies, 1987
- Learning translation invariant recognition in a massively parallel networks1987