IDEAS home Printed from https://ideas.repec.org/a/gam/jstats/v4y2021i4p51-892d673038.html
   My bibliography  Save this article

The One Standard Error Rule for Model Selection: Does It Work?

Author

Listed:
  • Yuchen Chen

    (Carlson School of Management, University of Minnesota, Minneapolis, MN 55455, USA)

  • Yuhong Yang

    (School of Statistics, University of Minnesota, Minneapolis, MN 55455, USA)

Abstract

Previous research provided a lot of discussion on the selection of regularization parameters when it comes to the application of regularization methods for high-dimensional regression. The popular “One Standard Error Rule” (1se rule) used with cross validation (CV) is to select the most parsimonious model whose prediction error is not much worse than the minimum CV error. This paper examines the validity of the 1se rule from a theoretical angle and also studies its estimation accuracy and performances in applications of regression estimation and variable selection, particularly for Lasso in a regression framework. Our theoretical result shows that when a regression procedure produces the regression estimator converging relatively fast to the true regression function, the standard error estimation formula in the 1se rule is justified asymptotically. The numerical results show the following: 1. the 1se rule in general does not necessarily provide a good estimation for the intended standard deviation of the cross validation error. The estimation bias can be 50–100% upwards or downwards in various situations; 2. the results tend to support that 1se rule usually outperforms the regular CV in sparse variable selection and alleviates the over-selection tendency of Lasso; 3. in regression estimation or prediction, the 1se rule often performs worse. In addition, comparisons are made over two real data sets: Boston Housing Prices (large sample size n , small/moderate number of variables p ) and Bardet–Biedl data (large p , small n ). Data guided simulations are done to provide insight on the relative performances of the 1se rule and the regular CV.

Suggested Citation

  • Yuchen Chen & Yuhong Yang, 2021. "The One Standard Error Rule for Model Selection: Does It Work?," Stats, MDPI, vol. 4(4), pages 1-25, November.
    • Handle: RePEc:gam:jstats:v:4:y:2021:i:4:p:51-892:d:673038
    as

    Download full text from publisher

    File URL: https://www.mdpi.com/2571-905X/4/4/51/pdf
    Download Restriction: no
    File URL: https://www.mdpi.com/2571-905X/4/4/51/
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Jiahua Chen & Zehua Chen, 2008. "Extended Bayesian information criteria for model selection with large model spaces," Biometrika, Biometrika Trust, vol. 95(3), pages 759-771.
    2. Harrison, David Jr. & Rubinfeld, Daniel L., 1978. "Hedonic housing prices and the demand for clean air," Journal of Environmental Economics and Management, Elsevier, vol. 5(1), pages 81-102, March.
    3. Wenjing Yang & Yuhong Yang, 2017. "Toward an objective and reproducible model choice via variable selection deviation," Biometrics, The International Biometric Society, vol. 73(1), pages 20-30, March.
    4. Zhang, Yongli & Yang, Yuhong, 2015. "Cross-validation for selecting a model selection procedure," Journal of Econometrics, Elsevier, vol. 187(1), pages 95-112.
    5. Chenglong Ye & Yi Yang & Yuhong Yang, 2018. "Sparsity Oriented Importance Learning for High-Dimensional Linear Regression," Journal of the American Statistical Association, Taylor & Francis Journals, vol. 113(524), pages 1797-1812, October.
    6. Hansheng Wang & Runze Li & Chih-Ling Tsai, 2007. "Tuning parameter selectors for the smoothly clipped absolute deviation method," Biometrika, Biometrika Trust, vol. 94(3), pages 553-568.
    Full references (including those not matched with items on IDEAS)

    Citations

    Citations are extracted by the CitEc Project, subscribe to its RSS feed for this item.
    as


    Cited by:

    1. Ervasti, Jenni & Pentti, Jaana & Seppälä, Piia & Ropponen, Annina & Virtanen, Marianna & Elovainio, Marko & Chandola, Tarani & Kivimäki, Mika & Airaksinen, Jaakko, 2023. "Prediction of bullying at work: A data-driven analysis of the Finnish public sector cohort study," Social Science & Medicine, Elsevier, vol. 317(C).

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Jie Ding & Vahid Tarokh & Yuhong Yang, 2018. "Model Selection Techniques -- An Overview," Papers 1810.09583, arXiv.org.
    2. Gaorong Li & Liugen Xue & Heng Lian, 2012. "SCAD-penalised generalised additive models with non-polynomial dimensionality," Journal of Nonparametric Statistics, Taylor & Francis Journals, vol. 24(3), pages 681-697.
    3. Tizheng Li & Xiaojuan Kang, 2022. "Variable selection of higher-order partially linear spatial autoregressive model with a diverging number of parameters," Statistical Papers, Springer, vol. 63(1), pages 243-285, February.
    4. Tang, Yanlin & Song, Xinyuan & Wang, Huixia Judy & Zhu, Zhongyi, 2013. "Variable selection in high-dimensional quantile varying coefficient models," Journal of Multivariate Analysis, Elsevier, vol. 122(C), pages 115-132.
    5. Yunxiao Chen & Xiaoou Li & Jingchen Liu & Zhiliang Ying, 2017. "Regularized Latent Class Analysis with Application in Cognitive Diagnosis," Psychometrika, Springer;The Psychometric Society, vol. 82(3), pages 660-692, September.
    6. Sakyajit Bhattacharya & Paul McNicholas, 2014. "A LASSO-penalized BIC for mixture model selection," Advances in Data Analysis and Classification, Springer;German Classification Society - Gesellschaft für Klassifikation (GfKl);Japanese Classification Society (JCS);Classification and Data Analysis Group of the Italian Statistical Society (CLADAG);International Federation of Classification Societies (IFCS), vol. 8(1), pages 45-61, March.
    7. Qingliang Fan & Yaqian Wu, 2020. "Endogenous Treatment Effect Estimation with some Invalid and Irrelevant Instruments," Papers 2006.14998, arXiv.org.
    8. Wei Qian & Yuhong Yang, 2013. "Model selection via standard error adjusted adaptive lasso," Annals of the Institute of Statistical Mathematics, Springer;The Institute of Statistical Mathematics, vol. 65(2), pages 295-318, April.
    9. Guo, Jie & Tang, Manlai & Tian, Maozai & Zhu, Kai, 2013. "Variable selection in high-dimensional partially linear additive models for composite quantile regression," Computational Statistics & Data Analysis, Elsevier, vol. 65(C), pages 56-67.
    10. Wang, Sheng & Zimmerman, Dale L. & Breheny, Patrick, 2020. "Sparsity-regularized skewness estimation for the multivariate skew normal and multivariate skew t distributions," Journal of Multivariate Analysis, Elsevier, vol. 179(C).
    11. Peng, Jingfu & Yang, Yuhong, 2022. "On improvability of model selection by model averaging," Journal of Econometrics, Elsevier, vol. 229(2), pages 246-262.
    12. Eun Ryung Lee & Hohsuk Noh & Byeong U. Park, 2014. "Model Selection via Bayesian Information Criterion for Quantile Regression Models," Journal of the American Statistical Association, Taylor & Francis Journals, vol. 109(505), pages 216-229, March.
    13. Xuan Liu & Jianbao Chen, 2021. "Variable Selection for the Spatial Autoregressive Model with Autoregressive Disturbances," Mathematics, MDPI, vol. 9(12), pages 1-20, June.
    14. Wenjing Yang & Yuhong Yang, 2017. "Toward an objective and reproducible model choice via variable selection deviation," Biometrics, The International Biometric Society, vol. 73(1), pages 20-30, March.
    15. Lin, Hongmei & Lian, Heng & Liang, Hua, 2019. "Rank reduction for high-dimensional generalized additive models," Journal of Multivariate Analysis, Elsevier, vol. 173(C), pages 672-684.
    16. Lian, Heng & Li, Jianbo & Tang, Xingyu, 2014. "SCAD-penalized regression in additive partially linear proportional hazards models with an ultra-high-dimensional linear part," Journal of Multivariate Analysis, Elsevier, vol. 125(C), pages 50-64.
    17. Lian, Heng, 2014. "Semiparametric Bayesian information criterion for model selection in ultra-high dimensional additive models," Journal of Multivariate Analysis, Elsevier, vol. 123(C), pages 304-310.
    18. Chen, Yunxiao & Li, Xiaoou & Liu, Jingchen & Ying, Zhiliang, 2017. "Regularized latent class analysis with application in cognitive diagnosis," LSE Research Online Documents on Economics 103182, London School of Economics and Political Science, LSE Library.
    19. Hu, Yuao & Lian, Heng, 2013. "Variable selection in a partially linear proportional hazards model with a diverging dimensionality," Statistics & Probability Letters, Elsevier, vol. 83(1), pages 61-69.
    20. Wang, Tao & Zhu, Lixing, 2011. "Consistent tuning parameter selection in high dimensional sparse linear regression," Journal of Multivariate Analysis, Elsevier, vol. 102(7), pages 1141-1151, August.

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gam:jstats:v:4:y:2021:i:4:p:51-892:d:673038. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: MDPI Indexing Manager (email available below). General contact details of provider: https://www.mdpi.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.