Evaluating feature selection for SVMs in high dimensions

Authors:
Roland Nilsson;José M. Peña;Johan Björkegren;Jesper Tegnér
Affiliations:
IFM Computational Biology, Linköping University, Linköping, Sweden;IFM Computational Biology, Linköping University, Linköping, Sweden;Gustav V Research Institute, Karolinska Institute, Stockholm, Sweden;IFM Computational Biology, Linköping University, Linköping, Sweden
Venue:
ECML'06 Proceedings of the 17th European conference on Machine Learning
Year:
2006

Citing 11
Cited 3

Support-Vector Networks

Machine Learning
Wrappers for feature subset selection

Artificial Intelligence - Special issue on relevance
Approximate statistical tests for comparing supervised classification learning algorithms

Neural Computation
Gene Selection for Cancer Classification using Support Vector Machines

Machine Learning
An introduction to variable and feature selection

The Journal of Machine Learning Research
Grafting: fast, incremental feature selection by gradient descent in function space

The Journal of Machine Learning Research
Use of the zero norm with linear models and kernel methods

The Journal of Machine Learning Research
A Feature Selection Newton Method for Support Vector Machine Classification

Computational Optimization and Applications
A comprehensive evaluation of multicategory classification methods for microarray gene expression cancer diagnosis

Bioinformatics
Editorial: The fundamental role of pattern recognition for gene-expression/microarray data in bioinformatics

Pattern Recognition
Efficient tuning of SVM hyperparameters using radius/margin bound and iterative algorithms

IEEE Transactions on Neural Networks

Incremental Bayesian Network Learning for Scalable Feature Selection

IDA '09 Proceedings of the 8th International Symposium on Intelligent Data Analysis: Advances in Intelligent Data Analysis VIII
Fuzzy-input fuzzy-output one-against-all support vector machines

KES'07/WIRN'07 Proceedings of the 11th international conference, KES 2007 and XVII Italian workshop on neural networks conference on Knowledge-based intelligent information and engineering systems: Part III
Training strategy of semantic concept detectors using support vector machine in naked image classification

PCM'10 Proceedings of the 11th Pacific Rim conference on Advances in multimedia information processing: Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

We perform a systematic evaluation of feature selection (FS) methods for support vector machines (SVMs) using simulated high- dimensional data (up to 5000 dimensions). Several findings previously reported at low dimensions do not apply in high dimensions. For example, none of the FS methods investigated improved SVM accuracy, indicating that the SVM built-in regularization is sufficient. These results were also validated using microarray data. Moreover, all FS methods tend to discard many relevant features. This is a problem for applications such as microarray data analysis, where identifying all biologically important features is a major objective.