Single-index quantile regression

JOURNAL OF MULTIVARIATE ANALYSIS

卷:101

Article

Wu, Tracy Z.²  Yu, Keming³  Yu, Yan¹ 

[1] Univ Cincinnati, Dept Quantitat Anal Operat Management, Cincinnati, OH 45221 USA

[2] JPMorgan Chase Bank, Consumer Risk Modeling & Analyt Grp, Columbus, OH 43240 USA

[3] Brunel Univ, Uxbridge UB8 3PH, Middx, England

关键词: Conditional quantile;    Dimension reduction;    Local polynomial smoothing;    Nonparametric model;    Semiparametric model;   

DOI  :  10.1016/j.jmva.2010.02.003

来源: Elsevier

【 摘 要 】

Nonparametric quantile regression with multivariate covariates is a difficult estimation problem due to the curse of dimensionality. To reduce the dimensionality while still retaining the flexibility of a nonparametric model, we propose modeling the conditional quantile by a single-index function g(0)(x(T)gamma(0),), where a univariate link function g(0)(.) is applied to a linear combination of covariates x(T)gamma(0), often called the single-index. We introduce a practical algorithm where the unknown link function g(0)(.) is estimated by local linear quantile regression and the parametric index is estimated through linear quantile regression. Large sample properties of estimators are studied, which facilitate further inference. Both the modeling and estimation approaches are demonstrated by simulation studies and real data applications. (C) 2010 Elsevier Inc. All rights reserved.

【 授权许可】

Free