A convex-Nonconvex strategy for grouped variable selection
From MaRDI portal
Abstract: This paper deals with the grouped variable selection problem. A widely used strategy is to augment the negative log-likelihood function with a sparsity-promoting penalty. Existing methods include the group Lasso, group SCAD, and group MCP. The group Lasso solves a convex optimization problem but is plagued by underestimation bias. The group SCAD and group MCP avoid this estimation bias but require solving a nonconvex optimization problem that may be plagued by suboptimal local optima. In this work, we propose an alternative method based on the generalized minimax concave (GMC) penalty, which is a folded concave penalty that maintains the convexity of the objective function. We develop a new method for grouped variable selection in linear regression, the group GMC, that generalizes the strategy of the original GMC estimator. We present an efficient algorithm for computing the group GMC estimator and also prove properties of the solution path to guide its numerical computation and tuning parameter selection in practice. We establish error bounds for both the group GMC and original GMC estimators. A rich set of simulation studies and a real data application indicate that the proposed group GMC approach outperforms existing methods in several different aspects under a wide array of scenarios.
Cites work
- A first-order primal-dual algorithm for convex problems with applications to imaging
- A general theory of concave regularization for high-dimensional sparse estimation problems
- A selective review of group selection in high-dimensional models
- A unified framework for high-dimensional analysis of M-estimators with decomposable regularizers
- Blockwise sparse regression
- Consistent group selection in high-dimensional linear regression
- Convex Analysis
- Convex analysis and monotone operator theory in Hilbert spaces
- Fast Nonconvex Nonsmooth Minimization Methods for Image Restoration and Reconstruction
- Group descent algorithms for nonconvex penalized linear and logistic regression models with grouped predictors
- High-dimensional statistics. A non-asymptotic viewpoint
- scientific article; zbMATH DE number 1547349 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- scientific article; zbMATH DE number 6438182 (Why is no real title available?)
- Hybrid safe-strong rules for efficient optimization in Lasso-type problems
- Linearly involved generalized Moreau enhanced models and their proximal splitting algorithm under overall convexity condition
- Model Selection and Estimation in Regression with Grouped Variables
- Nearly unbiased variable selection under minimax concave penalty
- Non-convex total variation regularization for convex denoising of signals
- On the Convergence of the Iterative Shrinkage/Thresholding Algorithm With a Weakly Convex Penalty
- Regularized \(M\)-estimators with nonconvexity: statistical and algorithmic theory for local optima
- Sparse Regularization via Convex Analysis
- Sparsity-Inducing Nonconvex Nonseparable Regularization for Convex Image Processing
- Strong oracle optimality of folded concave penalized estimation
- The composite absolute penalties family for grouped and hierarchical variable selection
- The Group Lasso for Logistic Regression
- The Lasso problem and uniqueness
- Variable selection in nonparametric varying-coefficient models for analysis of repeated measurements
- Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
Cited in
(4)- Group variable selection via a hierarchical lasso and its oracle property
- Concave group methods for variable selection and estimation in high-dimensional varying coefficient models
- Group variable selection via SCAD-L2
- Automatic selection by penalized asymmetric L q -norm in a high-dimensional model with grouped variables
This page was built for publication: A convex-Nonconvex strategy for grouped variable selection
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6144411)
