Abstract
A robust and sparse estimator for multinomial regression is proposed for high dimensional data. Robustness of the estimator is achieved by trimming the observations, and sparsity of the estimator is obtained by the elastic net penalty. In contrast to multi-group classifiers based on dimension reduction, this model is very appealing in terms of interpretation, since one obtains estimated coefficients individually for every group, and also the sparsity of the coefficients is group specific. Simulation studies are conducted to show the performance in comparison to the non-robust version of the multinomial regression estimator, and some real data examples underline the usefulness of this robust estimator particularly in terms of result interpretation and model diagnostics.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.References
Alfons A, Croux C, Gelper S (2013) Sparse least trimmed squares regression for analyzing high-dimensional large data sets. Ann Appl Stat 7(1):226–248
Castilla E, Ghosh A, Martin N, Pardo L (2018) New robust statistical procedures for polytomous logistic regression models. Biometrics 74(4):1282–1291
Cawley G, Talbot N, Girolami M (2006) Sparse multinomial logistic regression via bayesian L1 regularisation. Adv Neural Inf Process Syst 19:209–216
Clemmensen L, Hastie T, Witten D, Ersbøll B (2011) Sparse discriminant analysis. Technometrics 53(4):406–413
Development Core Team, R (2021) R Foundation for Statistical Computing Vienna Austria. https://www.R-project.org/
Friedman J, Hastie T, Tibshirani R (2010) Regularization paths for generalized linear models via coordinate descent. J Stat Softw 33(1):1–22
Friedman J, Hastie T, Tibshirani R, Narasimhan B, Tay K, Simon N, Qian J, Yang J (2021) glmnet: Lasso and Elastic-Net Regularized Generalized Linear Models. . R Foundation for Statistical Computing, Vienna, Austria. R package version 4.1–3. https://CRAN.R-project.org/package=glmnet
Hastie T, Buja A, Tibshirani R (1995) Penalized discriminant analysis. Ann Stat 23(1):73–102
Hubert M, Van Driessen K (2004) Fast and robust discriminant analysis. Comput Stat Data Anal 45(2):301–320
Johnson RA, Wichern DW (2007) Applied multivariate statistical analysis, 6th edn. Prentice Hall, Upper Saddle River
Kurnaz FS, Hoffmann I, Filzmoser P (2018) enetLTS: Robust and sparse estimation methods for high-dimensional linear and logistic regression. R Foundation for Statistical Computing, Vienna, Austria. R package https://CRAN.R-project.org/package=enetLTS
Kurnaz FS, Hoffmann I, Filzmoser P (2018) Robust and sparse estimation methods for high-dimensional linear and logistic regression. Chemom Intell Lab Syst 172:211–222
Lê Cao K, Boitard S, Besse P (2011) Sparse PLS discriminant analysis: biologically relevant feature selection and graphical displays for multiclass problems. BMC Bioinf 12:253
Matan O, Kiang R, Stenard C, Boser B, Denker J, Henderson D, Howard R, Hubbard W, Jackel L, Le Cun Y (1990) Handwritten character recognition using neural network architectures. In: Proceedings of the 4th US Postal Service Advanced Technology Conference 2:1003–1011
Ortner I, Filzmoser P, Croux C (2020) Robust and sparse multigroup classification by optimal scoring approach. Data Min Knowl Disc 34:723–741
Rousseeuw P, Leroy A (1987) Robust regression and outlier detection. Wiley Series in Probability and Statistics, John Wiley and Sons Ltd., New York
Rousseeuw PJ, Driessen KV (1999) A fast algorithm for the minimum covariance determinant estimator. Technometrics 41(3):212–223
Rousseeuw PJ, Driessen KV (2006) Computing its regression for large data sets. Data Min Knowl Disc 12(1):29–45
Tabatabai MA, Li H, Eby WM, Kengwoung-Keumo JJ, Manne U, Bae S, Fouad M, Singh KP (2014) Robust logistic and probit methods for binary and multinomial regression. J Biometrics Biostat 5(4)
Virta J, Koesner CL, Li B, Nordhausen K, Oja H, Radojicic U (2021) tensorbss: Blind source separation methods for tensor-valued observations. . R Foundation for Statistical Computing, Vienna, Austria. R package version 0.3.8 https://CRAN.R-project.org/package=tensorBSS
Yin M, Zeng D, Gao J, Wu Z, Xie S (2018) Robust multinomial logistic regression based on RPCA. IEEE J Sel Topics Signal Proces 12(6):1144–1154
Acknowledgements
This work is supported by grant TUBITAK 2219 from the Scientific and Technological Research Council of Turkey (TUBITAK).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The author declares that there is no conflict of interest.
Ethical approval
Not required ethics approval.
Informed Consent
Not applicable.
Additional information
Responsible editor: Johannes Fürnkranz.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kurnaz, F.S., Filzmoser, P. Robust and sparse multinomial regression in high dimensions. Data Min Knowl Disc 37, 1609–1629 (2023). https://doi.org/10.1007/s10618-023-00936-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10618-023-00936-6