Open Access Research Article

Quantification of the Impact of Feature Selection on the Variance of Cross-Validation Error Estimation

Yufei Xiao1*, Jianping Hua2 and Edward R Dougherty1,2

Author Affiliations

1 Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX 77843, USA

2 Computational Biology Division, Translational Genomics Research Institute, Phoenix, AZ 85004, USA

For all author emails, please log on.

EURASIP Journal on Bioinformatics and Systems Biology 2007, 2007:16354 doi:10.1155/2007/16354


The electronic version of this article is the complete one and can be found online at: http://bsb.eurasipjournals.com/content/2007/1/16354


Received:7 August 2006
Revisions received:21 December 2006
Accepted:26 December 2006
Published:19 February 2007

© 2007 Yufei Xiao et al.

This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Given the relatively small number of microarrays typically used in gene-expression-based classification, all of the data must be used to train a classifier and therefore the same training data is used for error estimation. The key issue regarding the quality of an error estimator in the context of small samples is its accuracy, and this is most directly analyzed via the deviation distribution of the estimator, this being the distribution of the difference between the estimated and true errors. Past studies indicate that given a prior set of features, cross-validation does not perform as well in this regard as some other training-data-based error estimators. The purpose of this study is to quantify the degree to which feature selection increases the variation of the deviation distribution in addition to the variation in the absence of feature selection. To this end, we propose the coefficient of relative increase in deviation dispersion (CRIDD), which gives the relative increase in the deviation-distribution variance using feature selection as opposed to using an optimal feature set without feature selection. The contribution of feature selection to the variance of the deviation distribution can be significant, contributing to over half of the variance in many of the cases studied. We consider linear-discriminant analysis, 3-nearest-neighbor, and linear support vector machines for classification; sequential forward selection, sequential forward floating selection, and the -test for feature selection; and -fold and leave-one-out cross-validation for error estimation. We apply these to three feature-label models and patient data from a breast cancer study. In sum, the cross-validation deviation distribution is significantly flatter when there is feature selection, compared with the case when cross-validation is performed on a given feature set. This is reflected by the observed positive values of the CRIDD, which is defined to quantify the contribution of feature selection towards the deviation variance.

Research Article

[12345678910111213]

References

  1. L Devroye, L Gyorfi, G Lugosi, A Probabilistic Theory of Pattern Recognition (Springer, New York, NY, USA, 1996)

  2. U Braga-Neto, ER Dougherty, Is cross-validation valid for small-sample microarray classification? Bioinformatics 20(3), 374–380 (2004). PubMed Abstract | Publisher Full Text OpenURL

  3. U Braga-Neto, ER Dougherty, Bolstered error estimation. Pattern Recognition 37(6), 1267–1281 (2004). Publisher Full Text OpenURL

  4. C Sima, U Braga-Neto, ER Dougherty, Superior feature-set ranking for small samples using bolstered error estimation. Bioinformatics 21(7), 1046–1054 (2005). PubMed Abstract | Publisher Full Text OpenURL

  5. C Sima, S Attoor, U Brag-Neto, J Lowey, E Suh, ER Dougherty, Impact of error estimation on feature selection. Pattern Recognition 38(12), 2472–2482 (2005). Publisher Full Text OpenURL

  6. AM Molinaro, R Simon, RM Pfeiffer, Prediction error estimation: a comparison of resampling methods. Bioinformatics 21(15), 3301–3307 (2005). PubMed Abstract | Publisher Full Text OpenURL

  7. P Pudil, J Novovicova, J Kittler, Floating search methods in feature selection. Pattern Recognition Letters 15(11), 1119–1125 (1994). Publisher Full Text OpenURL

  8. Y Xiao, J Hua, ER Dougherty, Feature selection increases cross-validation imprecision. Proceedings of the 4th IEEE International Workshop on Genomic Signal Processing and Statistics (GENSIPS '06), College Station, Tex, USA, May 2006

  9. LJ van't Veer, H Dai, MJ van de Vijver, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 415(6871), 530–536 (2002). PubMed Abstract | Publisher Full Text OpenURL

  10. MJ van de Vijver, YD He, LJ van't Veer, et al. A gene-expression signature as a predictor of survival in breast cancer. New England Journal of Medicine 347(25), 1999–2009 (2002). PubMed Abstract | Publisher Full Text OpenURL

  11. A Choudhary, M Brun, J Hua, J Lowey, E Suh, ER Dougherty, Genetic test bed for feature selection. Bioinformatics 22(7), 837–842 (2006). PubMed Abstract | Publisher Full Text OpenURL

  12. A Jain, D Zongker, Feature selection: evaluation, application, and small sample performance. IEEE Transactions on Pattern Analysis and Machine Intelligence 19(2), 153–158 (1997). Publisher Full Text OpenURL

  13. M Kudo, J Sklansky, Comparison of algorithms that select features for pattern classifiers. Pattern Recognition 33(1), 25–41 (2000). Publisher Full Text OpenURL