×

Semiparametric estimation for linear regression with symmetric errors. (English) Zbl 1510.62292

Summary: To avoid the effect of distributional misspecification in the model-based regression, we propose an essentially nonparametric symmetric error distribution and construct a so-called doubly smoothed (DS) likelihood function by applying the same amount of smoothing to both the model and given data. To compute the DS maximum likelihood estimator based on the DS likelihood, we propose an approximated DS likelihood which has the form of a semiparametric mixture likelihood and apply some existing algorithms in the nonparametric mixture literature. The consistency of the DS maximum likelihood estimator is also established with any fixed smoothing parameter. Through numerical studies, we demonstrate that the proposed regression coefficient estimator has relatively good performance in terms of efficiency across a wide range of error distributions and robustness against outliers.

MSC:

62J05 Linear regression; mixed models
62G05 Nonparametric estimation
62F35 Robustness and adaptive procedures (parametric inference)
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Andersen, R., (Modern Methods for Robust Regression. Modern Methods for Robust Regression, Quantitative Applications in the Social Sciences, vol. 152 (2008), SAGE Publications: SAGE Publications Thousand Oaks)
[2] Baltagi, B. H., Econometrics (2011), Springer: Springer Berlin · Zbl 1216.62188
[3] Bartolucci, F.; Scaccia, L., The use of mixtures for dealing with non-normal regression errors, Comput. Statist. Data Anal., 48, 821-834 (2005) · Zbl 1429.62284
[4] Bickel, P. J., On adaptive estimation, Ann. Statist., 10, 647-671 (1982) · Zbl 0489.62033
[5] Bucevska, V., Heteroscedasticity, (Lovric, M., International Encyclopedia of Statistical Science (2011), Springer: Springer Berlin), 630-633 · Zbl 1241.62001
[6] Chae, M.; Kim, Y.; Kleijn, B. J.K., The semi-parametric Bernstein-von Mises theorem for regression models with symmetric errors, Statist. Sinica, 29, 1465-1487 (2019) · Zbl 1432.62202
[7] Charnigo, R. J.; Pilla, R. S., Semiparametric mixtures of generalized exponential families, Scand. J. Stat., 34, 535-551 (2007) · Zbl 1150.62017
[8] Chatterjee, S.; Hadi, A. S., Regression Analysis By Example (2012), Wiley: Wiley Hoboken · Zbl 1263.62099
[9] Chee, C.-S.; Wang, Y., Estimation of finite mixtures with symmetric components, Stat. Comput., 23, 233-249 (2013) · Zbl 1322.62013
[10] Jaki, T.; West, R. W., Maximum kernel likelihood estimation, J. Comput. Graph. Statist., 17, 976-993 (2008)
[11] Koenker, R. W.; Bassett, G. W., Robust tests for heteroscedasticity based on regression quantiles, Econometrica, 50, 43-61 (1982) · Zbl 0482.62023
[12] Lindsay, B. G., (Mixture Models: Theory, Geometry and Applications. Mixture Models: Theory, Geometry and Applications, NSF-CBMS Regional Conference Series in Probability and Statistics, vol. 5 (1995), Institute of Mathematical Statistics: Institute of Mathematical Statistics Hayward) · Zbl 1163.62326
[13] Lindsay, B. G.; Lesperance, M. L., A review of semiparametric mixture models, J. Statist. Plann. Inference, 47, 29-39 (1995) · Zbl 0832.62027
[14] Martin, R.; Han, Z., A semiparametric scale-mixture regression model and predictive recursion maximum likelihood, Comput. Statist. Data Anal., 94, 75-85 (2016) · Zbl 1468.62136
[15] McDonald, J. B.; Newey, W. K., Partially adaptive estimation of regression models via the generalized t distribution, Econometric Theory, 4, 428-457 (1988)
[16] Roeder, K. M., Density estimation with confidence sets exemplified by superclusters and voids in the galaxies, J. Amer. Statist. Assoc., 85, 617-624 (1990) · Zbl 0704.62103
[17] Seo, B., The doubly smoothed maximum likelihood estimation for location-shifted semiparametric mixtures, Comput. Statist. Data Anal., 108, 27-39 (2017) · Zbl 1466.62191
[18] Seo, B.; Lindsay, B. G., A universally consistent modification of maximum likelihood, Statist. Sinica, 23, 467-487 (2013) · Zbl 1284.62233
[19] Seo, B.; Noh, J.; Lee, T.; Yoon, Y. J., Adaptive robust regression with continuous gaussian scale mixture errors, J. Korean Stat. Soc., 46, 113-125 (2017) · Zbl 1357.62246
[20] Usta, I.; Kantar, Y. M., On the performance of the flexible maximum entropy distributions within partially adaptive estimation, Comput. Statist. Data Anal., 55, 2172-2182 (2011) · Zbl 1328.62447
[21] Wang, Y., On fast computation of the non-parametric maximum likelihood estimate of a mixing distribution, J. R. Stat. Soc. Ser. B Stat. Methodol., 69, 185-198 (2007) · Zbl 1120.62022
[22] Wang, Y., Maximum likelihood computation for fitting semiparametric mixture models, Stat. Comput., 20, 75-86 (2010)
[23] Wang, Y.; Chee, C.-S., Density estimation using non-parametric and semi-parametric mixtures, Stat. Model., 12, 67-92 (2012) · Zbl 1420.62164
[24] Wilcox, R. R., Introduction to Robust Estimation and Hypothesis Testing (2017), Academic Press: Academic Press San Diego · Zbl 1362.62003
[25] Yao, W.; Zhao, Z., Kernel density-based linear regression estimate, Comm. Statist. Theory Methods, 42, 4499-4512 (2013) · Zbl 1279.62088
[26] Yu, C.; Yao, W., Robust linear regression: A review and comparison, Comm. Statist. Simulation Comput., 46, 6261-6282 (2017) · Zbl 1388.62070
[27] Yuan, A.; De Gooijer, J. G., Semiparametric regression with kernel error model, Scand. J. Stat., 34, 841-869 (2007) · Zbl 1157.62045
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.