Efficient quantile regression for heteroscedastic models

Yoonsuh Jung, Yoonkyung Lee, Steven N. MacEachern

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Quantile regression (QR) provides estimates of a range of conditional quantiles. This stands in contrast to traditional regression techniques, which focus on a single conditional mean function. Lee et al. [Regularization of case-specific parameters for robustness and efficiency. Statist Sci. 2012;27(3):350–372] proposed efficient QR by rounding the sharp corner of the loss. The main modification generally involves an asymmetric ℓ2 adjustment of the loss function around zero. We extend the idea of ℓ2 adjusted QR to linear heterogeneous models. The ℓ2 adjustment is constructed to diminish as sample size grows. Conditions to retain consistency properties are also provided.

Original languageEnglish
Pages (from-to)2548-2568
Number of pages21
JournalJournal of Statistical Computation and Simulation
Volume85
Issue number13
DOIs
Publication statusPublished - 2015 Sep 2
Externally publishedYes

Fingerprint

Heteroscedastic Model
Quantile Regression
Adjustment
Conditional Quantiles
Regression Estimate
Rounding
Loss Function
Regularization
Sample Size
Regression
Robustness
Zero
Range of data
Quantile regression
Model

Keywords

  • check loss function
  • heteroscedasticity
  • quantile regression

ASJC Scopus subject areas

  • Statistics and Probability
  • Modelling and Simulation
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Efficient quantile regression for heteroscedastic models. / Jung, Yoonsuh; Lee, Yoonkyung; MacEachern, Steven N.

In: Journal of Statistical Computation and Simulation, Vol. 85, No. 13, 02.09.2015, p. 2548-2568.

Research output: Contribution to journalArticle

Jung, Yoonsuh ; Lee, Yoonkyung ; MacEachern, Steven N. / Efficient quantile regression for heteroscedastic models. In: Journal of Statistical Computation and Simulation. 2015 ; Vol. 85, No. 13. pp. 2548-2568.
@article{bd451fc4587f4ecfbcddf810451464a2,
title = "Efficient quantile regression for heteroscedastic models",
abstract = "Quantile regression (QR) provides estimates of a range of conditional quantiles. This stands in contrast to traditional regression techniques, which focus on a single conditional mean function. Lee et al. [Regularization of case-specific parameters for robustness and efficiency. Statist Sci. 2012;27(3):350–372] proposed efficient QR by rounding the sharp corner of the loss. The main modification generally involves an asymmetric ℓ2 adjustment of the loss function around zero. We extend the idea of ℓ2 adjusted QR to linear heterogeneous models. The ℓ2 adjustment is constructed to diminish as sample size grows. Conditions to retain consistency properties are also provided.",
keywords = "check loss function, heteroscedasticity, quantile regression",
author = "Yoonsuh Jung and Yoonkyung Lee and MacEachern, {Steven N.}",
year = "2015",
month = "9",
day = "2",
doi = "10.1080/00949655.2014.967244",
language = "English",
volume = "85",
pages = "2548--2568",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "13",

}

TY - JOUR

T1 - Efficient quantile regression for heteroscedastic models

AU - Jung, Yoonsuh

AU - Lee, Yoonkyung

AU - MacEachern, Steven N.

PY - 2015/9/2

Y1 - 2015/9/2

N2 - Quantile regression (QR) provides estimates of a range of conditional quantiles. This stands in contrast to traditional regression techniques, which focus on a single conditional mean function. Lee et al. [Regularization of case-specific parameters for robustness and efficiency. Statist Sci. 2012;27(3):350–372] proposed efficient QR by rounding the sharp corner of the loss. The main modification generally involves an asymmetric ℓ2 adjustment of the loss function around zero. We extend the idea of ℓ2 adjusted QR to linear heterogeneous models. The ℓ2 adjustment is constructed to diminish as sample size grows. Conditions to retain consistency properties are also provided.

AB - Quantile regression (QR) provides estimates of a range of conditional quantiles. This stands in contrast to traditional regression techniques, which focus on a single conditional mean function. Lee et al. [Regularization of case-specific parameters for robustness and efficiency. Statist Sci. 2012;27(3):350–372] proposed efficient QR by rounding the sharp corner of the loss. The main modification generally involves an asymmetric ℓ2 adjustment of the loss function around zero. We extend the idea of ℓ2 adjusted QR to linear heterogeneous models. The ℓ2 adjustment is constructed to diminish as sample size grows. Conditions to retain consistency properties are also provided.

KW - check loss function

KW - heteroscedasticity

KW - quantile regression

UR - http://www.scopus.com/inward/record.url?scp=84930574539&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84930574539&partnerID=8YFLogxK

U2 - 10.1080/00949655.2014.967244

DO - 10.1080/00949655.2014.967244

M3 - Article

VL - 85

SP - 2548

EP - 2568

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 13

ER -