On varying-coefficient independence screening for high-dimensional varying-coefficient models
- 1 January 2014
- journal article
- research article
- Published by Statistica Sinica (Institute of Statistical Science) in Statistica Sinica
- Vol. 24 (4) , 1735-1752
Abstract
Varying coefficient models have been widely used in longitudinal data analysis, nonlinear time series, survival analysis, and so on. They are natural non-parametric extensions of the classical linear models in many contexts, keeping good interpretability and allowing us to explore the dynamic nature of the model. Recently, penalized estimators have been used for fitting varying-coefficient models for high-dimensional data. In this paper, we propose a new computationally attractive algorithm called IVIS for fitting varying-coefficient models in ultra-high dimensions. The algorithm first fits a gSCAD penalized varying-coefficient model using a subset of covariates selected by a new varying-coefficient independence screening (VIS) technique. The sure screening property is established for VIS. The proposed algorithm then iterates between a greedy conditional VIS step and a gSCAD penalized fitting step. Simulation and a real data analysis demonstrate that IVIS has very competitive performance for moderate sample size and high dimension.Keywords
This publication has 13 references indexed in Scilit:
- Robust rank correlation based screeningThe Annals of Statistics, 2012
- Model-Free Feature Screening for Ultrahigh-Dimensional DataJournal of the American Statistical Association, 2011
- Nonparametric Independence Screening in Sparse Ultra-High-Dimensional Additive ModelsJournal of the American Statistical Association, 2011
- Simultaneous analysis of Lasso and Dantzig selectorThe Annals of Statistics, 2009
- Shrinkage Estimation of the Varying Coefficient ModelJournal of the American Statistical Association, 2009
- Lasso-type recovery of sparse representations for high-dimensional dataThe Annals of Statistics, 2009
- Variable Selection in Nonparametric Varying-Coefficient Models for Analysis of Repeated MeasurementsJournal of the American Statistical Association, 2008
- High-dimensional generalized linear models and the lassoThe Annals of Statistics, 2008
- The Dantzig selector: Statistical estimation when p is much larger than nThe Annals of Statistics, 2007
- Group SCAD regression analysis for microarray time course gene expression dataBioinformatics, 2007