Weight choice for penalized composite quantile regression and for model averaging

Jing Zhou; Gerda Claeskens; Daumantas Bloznelis

Weight choice for penalized composite quantile regression and for model averaging

Jing Zhou
, Gerda Claeskens
, Daumantas Bloznelis

Statistics

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

Abstract

For composite estimation, weights are needed to combine the different loss functions; likewise, for model averaging, weights are used to average the
different estimators. We investigate the choice of weights for both types of estimators in the setting of quantile regression. We pay particular attention to the high-dimensional case where due to the regularization, different expressions may
be used to define so-called optimal weights

Original language	English
Title of host publication	Proceedings of the 33rd International Workshop on Statistical Modelling
Publisher	Statistical Modelling Society
Pages	219-224
Number of pages	6
Volume	2
Publication status	Published - 16 Jul 2018
Event	33rd International Workshop on Statistical Modelling - Bristol, United Kingdom Duration: 16 Jul 2018 → 20 Jul 2018

Conference

Conference	33rd International Workshop on Statistical Modelling
Country/Territory	United Kingdom
City	Bristol
Period	16/07/18 → 20/07/18

Keywords

quantile regression
penalized estimation
composite estimation
weight choice

Cite this

@inproceedings{4647c654fff5450a8dcfcaabd63bf07e,

title = "Weight choice for penalized composite quantile regression and for model averaging",

abstract = "For composite estimation, weights are needed to combine the different loss functions; likewise, for model averaging, weights are used to average the different estimators. We investigate the choice of weights for both types of estimators in the setting of quantile regression. We pay particular attention to the high-dimensional case where due to the regularization, different expressions may be used to define so-called optimal weights",

keywords = "quantile regression, penalized estimation, composite estimation, weight choice",

author = "Jing Zhou and Gerda Claeskens and Daumantas Bloznelis",

year = "2018",

month = jul,

day = "16",

language = "English",

volume = "2",

pages = "219--224",

booktitle = "Proceedings of the 33rd International Workshop on Statistical Modelling",

publisher = "Statistical Modelling Society",

address = "Netherlands",

note = "33rd International Workshop on Statistical Modelling ; Conference date: 16-07-2018 Through 20-07-2018",

}

TY - GEN

T1 - Weight choice for penalized composite quantile regression and for model averaging

AU - Zhou, Jing

AU - Claeskens, Gerda

AU - Bloznelis, Daumantas

PY - 2018/7/16

Y1 - 2018/7/16

N2 - For composite estimation, weights are needed to combine the different loss functions; likewise, for model averaging, weights are used to average the different estimators. We investigate the choice of weights for both types of estimators in the setting of quantile regression. We pay particular attention to the high-dimensional case where due to the regularization, different expressions may be used to define so-called optimal weights

AB - For composite estimation, weights are needed to combine the different loss functions; likewise, for model averaging, weights are used to average the different estimators. We investigate the choice of weights for both types of estimators in the setting of quantile regression. We pay particular attention to the high-dimensional case where due to the regularization, different expressions may be used to define so-called optimal weights

KW - quantile regression

KW - penalized estimation

KW - composite estimation

KW - weight choice

M3 - Conference contribution

VL - 2

SP - 219

EP - 224

BT - Proceedings of the 33rd International Workshop on Statistical Modelling

PB - Statistical Modelling Society

T2 - 33rd International Workshop on Statistical Modelling

Y2 - 16 July 2018 through 20 July 2018

ER -

Weight choice for penalized composite quantile regression and for model averaging

Abstract

Conference

Keywords

Fingerprint

Cite this