Functional sparsity: global versus local
Haonan Wang, Bo Kai · Statistica Sinica · 2014
We consider the model selection problem in nonparametric regression. The notion of functional sparsity is a generalization of parameter sparsity in parametric models. In particular, two different types of sparsity are studied, including both global sparsity and local sparsity. The goal is to produce a sparse estimate which assigns zero values over regions where the true underlying function is zero, i.e., no relationship between the response variable and the predictor variable. Most classical smoothing techniques yield consistent estimates with no sparsity. Here, a penalized least squares procedure, based on the basis function approximation and the group bridge penalty function, is proposed for simultaneous function estimation and zero subregion detection. Asymptotic properties, including both consistency in estimation and sparsistency in model selection, of the procedure are established. The methodology is illustrated through simulation studies and a case study.
Cited by 13 works. doi:10.5705/ss.2014.035
Authors
References
- A Practical Guide to Splines
- Regression Shrinkage and Selection Via the Lasso
- Least angle regression
- Spline Models for Observational Data.
- Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
- Model Selection and Estimation in Regression with Grouped Variables
- Applied nonparametric regression
- Spline Functions: Basic Theory.
- The Adaptive Lasso and Its Oracle Properties
- Optimal Global Rates of Convergence for Nonparametric Regression