Abstract
Classification and regression problems in which the number of predictor variables is larger than the number of observations are increasingly common with rapid technological advances in data collection. Because some of these variables may have little or no influence on the response, methods that can identify the unimportant variables are needed. Two methods that have been proposed for this purpose are EARTH and Random forest (RF). This article presents an alternative method, derived from the GUIDE classification and regression tree algorithm, that employs recursive partitioning to determine the degree of importance of the variables. Simulation experiments show that the new method improves the prediction accuracy of several nonparametric regression models more than Random forest and EARTH. The results indicate that it is not essential to correctly identify all the important variables in every situation. Conditions for which this occurs are obtained for the linear model. The article concludes with an application of the new method to identify rare molecules in a large genomic data set.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Breiman L (2001) Random forests. Mach Learn 45:5–32
Breiman L, Friedman JH, Olshen RA, Stone CJ (1984) Classification and regression trees. Wadsworth, Belmont
Chernoff H, Lo S-H, Zheng T (2009) Discovering influential variables: a method of partitions. Ann Appl Stat 3:1335–1369
Doksum K, Tang S, Tsui K-W (2008) Nonparametric variable selection: the EARTH algorithm. J Am Stat Assoc 103:1609–1620
Friedman J (1991) Multivariate adaptive regression splines (with discussion). Ann Stat 19:1–141
Loh W-Y (2002) Regression trees with unbiased variable selection and interaction detection. Stat Sin 12:361–386
Loh W-Y (2009) Improving the precision of classification trees. Ann Appl Stat 3:1710–1737
Satterthwaite FE (1946) An approximate distribution of estimates of variance components. Biometrics Bull 2:110–114
Seber GAF, Lee AJ (2003) Linear regression analysis. 2nd edn. Wiley, New York
Strobl C, Boulesteix A, Zeileis A, Hothorn T (2007) Bias in random forest variable importance measures: illustrations, sources and a solution. BMC Bioinf 8:25
Tuv E, Borisov A, Torkkola K (2006) Feature selection using ensemble based ranking against artificial contrasts. In: IJCNN ’06. International joint conference on neural networks, Vancouver, Canada
Acknowledgements
This research was partially supported by the U.S. Army Research Office under grants W911NF-05-1-0047 and W911NF-09-1-0205. The author is grateful to K. Doksum, S. Tang, and K. Tsui for helpful discussions and to S. Tang for the computer code for EARTH.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer Science+Business Media, LLC
About this paper
Cite this paper
Loh, WY. (2012). Variable Selection for Classification and Regression in Large p, Small n Problems. In: Barbour, A., Chan, H., Siegmund, D. (eds) Probability Approximations and Beyond. Lecture Notes in Statistics(), vol 205. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-1966-2_10
Download citation
DOI: https://doi.org/10.1007/978-1-4614-1966-2_10
Published:
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-1965-5
Online ISBN: 978-1-4614-1966-2
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)