×

Quantile regression: A nonparametric approach. (English) Zbl 0726.62057

Summary: Regression on any p-th quantile is considered through nonparametric modelling. The nonparametric technique used is moving parabolic fit which is known to be adaptive and to reduce bias in the usual mean regression. The quantile problem reduces to solving weighted linear regression in \(L_ 1\) norm at each x-point and the iteratively reweighted least squares algorithm is particularly suitable for this. Convergence of the IRLS algorithm is shown from a broad perspective. Implementation details are given and some illustrations are presented.

MSC:

62G07 Density estimation
62J05 Linear regression; mixed models
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Basset, G.; Koenker, R.: An empirial quantile function for linear models with iid errors. J. amer. Stat. assoc. 77, 407-415 (1982) · Zbl 0493.62047
[2] Cleveland, W. S.: Robust locally weighted regression and smoothing scatterplots. J. amer. Stat. assoc. 74, 829-836 (1979) · Zbl 0423.62029
[3] Ekblom, H.: The L1-estimate as limiting case of an lp- or huber-estimate. Statistical data analysis based on the L1 norm and related methods, 109-116 (1987)
[4] Huber, P. J.: Robust statistics. (1981) · Zbl 0536.62025
[5] Lejeune, M.: Optimization in nonparametric regression. Compstat 84, Proceedings in computational statistics, 412-426 (1984)
[6] Lejeune, M.: Estimation non paramétrique par noyaux: régression polynômiale mobile. Revue de statistique appliquée 33, No. 3, 43-67 (1985) · Zbl 0569.62025
[7] Narula, S. C.: The minimum sum of absolute errors regression. J. quality tech. 19, 37-45 (1987)
[8] Nchs: NCHS growth charts, 1976. Vital and health statistics (1976)
[9] Portnoy, S.: Using regression fractiles to identify outliers. Statistical data analysis based on the L1 norm and related methods, 345-356 (1987)
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.