Robust variable selection via the weighted elastic-net multi-step screening procedure

Year 2025, Volume: 54 Issue: 6, 2350 - 2362, 30.12.2025

Yunlu Jıang , Huijie Lu , Xiaowen Huang , Ruizhe Jiang

https://doi.org/10.15672/hujms.1701976

Abstract

Variable selection in high-dimensional data remains a critically important, yet challenging task, particularly when confronted with highly correlated features and outliers. In this paper, we propose a novel robust variable selection method via the weighted elastic-net multi-step screening procedure. Our proposed method is not only robust to heavy-tailed errors or high leverage points, but can also handle highly correlated covariates and high-dimensional data sets with $p>n$, where $p$ is the number of predictors and $n$ is the sample size. In addition, a multi-step iterative algorithm is introduced to obtain the proposed estimator. Finally, extensive numerical simulations and a real-world NASDAQ index tracking application are conducted to illustrate the merits of the proposed method. The results indicate that our proposed method has a better finite-sample performance than some existing methods when there exist highly correlated covariates and outliers in the high-dimensional linear regression model.

Keywords

high-dimensional data , multicollinearity , robustness , weighted elastic-net multi-step screening

Ethical Statement

The authors declare no conflict of interest related to this study.

Project Number

This work is supported by the National Science Foundation of China (12171203), and the Fundamental Research Funds for the Central Universities (23JNQMX21)

Thanks

The authors would like to express their heartfelt gratitude to the reviewers for their valuable feedback and insightful suggestions, which have greatly enhanced the quality of this work.

References

• [1] A. Alfons, C. Croux and S. Gelper, Sparse least trimmed squares regression for analyzing high-dimensional large data sets, Ann. Appl. Stat., 226-248, 2013.
• [2] O. Arslan, Weighted LAD-LASSO method for robust parameter estimation and variable selection in regression, Comput. Stat. Data. An. 56 (6), 1952-1965, 2012.
• [3] K. Boudt, P.J. Rousseeuw, S. Vanduffel and T. Verdonck, The minimum regularized covariance determinant estimator, Stat. Comput. 30 (1), 113-128, 2020.
• [4] J. Bradic, J. Fan and W. Wang, Penalized composite quasi-likelihood for ultrahigh dimensional variable selection, J. R. Stat. Soc. Ser. B Stat. Methodol. 73 (3), 325- 349, 2011.
• [5] P. Bühlmann, M. Kalisch and M.H. Maathuis, Variable selection in high-dimensional linear models: partially faithful distributions and the PC-simple algorithm, Biometrika 97 (2), 261-278, 2010.
• [6] X. Chen, Z.J. Wang and M.J. McKeown, Asymptotic analysis of robust LASSOs in the presence of noise with large variance, IEEE Trans. Inf. Theory 56 (10), 5131-5149, 2010.
• [7] H. Cho and P. Fryzlewicz, High dimensional variable selection via tilting, J. R. Stat. Soc. Ser. B Stat. Methodol. 74 (3), 593-622, 2012.
• [8] J. Fan, Y. Fan and E. Barut, Adaptive robust variable selection, Ann. Stat. 42 (1), 324, 2014.
• [9] J. Fan and R. Li, Nonconcave penalized likelihood with NP-dimensionality, IEEE. Trans. Inf. Theory. 57 (8), 5467-5484, 2011.
• [10] J. Fan and R. Li, Variable selection via nonconcave penalized likelihood and its oracle properties, J. Amer. Stat. Assoc. 96 (456), 1348-1360, 2001.
• [11] Y. Jiang, Y. Wang, J. Jiang, B. Xie, J. Liao and W. Liao, Outlier detection and robust variable selection via the penalized weighted LAD-LASSO method, J. Appl. Stat. 48 (2), 234-246, 2021.
• [12] Y. Jiang, Y.G. Wang, L. Fu and X. Wang, Robust estimation using modified Huber’s functions with new tails, Technometrics 61 (1), 111-122, 2019.
• [13] B.A. Johnson and L. Peng, Rank-based variable selection, J. Nonparam. Stat. 20 (3), 241-252, 2008.
• [14] R.J. Karunamuni, L. Kong and W. Tu, Efficient robust doubly adaptive regularized regression with applications, Stat. Methods Med. Res. 28 (7), 2210-2226, 2019.
• [15] C. Leng, Variable selection and coefficient estimation via regularized rank regression, Stat. Sin., 167-181, 2010.
• [16] N. Li, Efficient sparse portfolios based on composite quantile regression for highdimensional index tracking, J. Stat. Comput. Simul. 90 (8), 1466-1478, 2020.
• [17] G. Li, H. Peng and L. Zhu, Nonconcave penalized M-estimation with a diverging number of parameters, Stat. Sin., 391-419, 2011.
• [18] P. Rousseeuw and K. Van Driessen, A fast algorithm for the minimum covariance determinant estimator, Technometrics 42 (3), 212-223, 1999.
• [19] V. Ročková and E.I. George, The spike-and-slab lasso, J. Am. Stat. Assoc. 113 (512), 431-444, 2018.
• [20] R. Tibshirani, Regression shrinkage and selection via the lasso, J. R. Stat. Soc. Ser. B Stat. Methodol. 58 (1), 267-288, 1996.
• [21] X. Wang, Y. Jiang, M. Huang and H. Zhang, Robust variable selection with exponential squared loss, J. Am. Stat. Assoc. 108 (502), 632-643, 2013.
• [22] H. Wang, G. Li and G. Jiang, Robust regression shrinkage and consistent variable selection through the LAD-Lasso, J. Bus. Econ. Stat. 25 (3), 347-355, 2007.
• [23] L. Wang and R. Li, Weighted Wilcoxon-type smoothly clipped absolute deviation method, Biometrics 65 (2), 564-571, 2009.
• [24] Y. Wu and Y. Liu, Variable selection in quantile regression, Stat. Sin., 801-817, 2009.
• [25] C. Wen, X. Wang and S. Wang, Laplace error penalty-based variable selection in high dimension, Scand. J. Stat., 42 (3), 685-700, 2015.
• [26] Y. Xue, J. Ren and B. Yang, Enmsp: an elastic-net multi-step screening procedure for high-dimensional regression, Stat. Comput. 34 (2), 79, 2024.
• [27] M. Yuan and Y. Lin, Model selection and estimation in regression with grouped variables, J. R. Stat. Soc. Ser. B Stat. Methodol. 68 (1), 49-67, 2006.
• [28] Z. Yang, L. Fu, Y.G. Wang, Z. Dong and Y. Jiang, A robust and efficient variable selection method for linear regression, J. App. Stat. 49 (14), 3677-3692, 2022.
• [29] C.H. Zhang, Nearly unbiased variable selection under minimax concave penalty, Ann. Stat. 38 (2), 894-942, 2010.
• [30] T. Zhang, Analysis of multi-stage convex relaxation for sparse regularization, J. Mach. Learn. Res. 11 (3), 2010.
• [31] J. Zhu, X. Wang, L. Hu, J. Huang, K. Jiang, Y. Zhang, S. Lin and J. Zhu, abess: A fast best-subset selection library in python and r, J. Mach. Learn. Res. 23 (202), 1-7, 2022.
• [32] J. Zhu, C. Wen, J. Zhu, H. Zhang and X. Wang, A polynomial algorithm for bestsubset selection problem, Proc. Natl. Acad. Sci. 117 (51), 33117-33123, 2020.
• [33] H. Zou, The adaptive lasso and its oracle properties, J. Am. Stat. Assoc. 101 (476), 1418-1429, 2006.
• [34] H. Zou and T. Hastie, Regularization and variable selection via the elastic net, J. R. Stat. Soc. Ser. B Stat. Methodol. 67 (2), 301-320, 2005.
• [35] H. Zou and M. Yuan, Composite quantile regression and the oracle model selection theory, Ann. Stat. 36 (3), 1108-1126, 2008.
• [36] H. Zou and H.H. Zhang, On the adaptive elastic-net with a diverging number of parameters, Ann. Stat. 37 (4), 1733, 2009.