Structured sparsity via alternating direction methods
From MaRDI portal
Abstract: We consider a class of sparse learning problems in high dimensional feature space regularized by a structured sparsity-inducing norm which incorporates prior knowledge of the group structure of the features. Such problems often pose a considerable challenge to optimization algorithms due to the non-smoothness and non-separability of the regularization term. In this paper, we focus on two commonly adopted sparsity-inducing regularization terms, the overlapping Group Lasso penalty -norm and the -norm. We propose a unified framework based on the augmented Lagrangian method, under which problems with both types of regularization and their variants can be efficiently solved. As the core building-block of this framework, we develop new algorithms using an alternating partial-linearization/splitting technique, and we prove that the accelerated versions of these algorithms require iterations to obtain an -optimal solution. To demonstrate the efficiency and relevance of our algorithms, we test them on a collection of data sets and apply them to two real-world problems to compare the relative merits of the two norms.
Recommendations
Cited in
(28)- scientific article; zbMATH DE number 7403729 (Why is no real title available?)
- An FFT-based fast gradient method for elastic and inelastic unit cell homogenization problems
- Two-layer feature reduction for sparse-group Lasso via decomposition of convex sets
- scientific article; zbMATH DE number 6253954 (Why is no real title available?)
- Fast algorithms for structured sparsity (ICALP 2015 invited tutorial)
- [HDDA] sparse subspace constrained partial least squares
- Weakly decomposable regularization penalties and structured sparsity
- Doubly majorized algorithm for sparsity-inducing optimization problems with regularizer-compatible constraints
- Boosting with structural sparsity: a differential inclusion approach
- Learning the Structure for Structured Sparsity
- Iteratively reweighted group Lasso based on log-composite regularization
- scientific article; zbMATH DE number 7370569 (Why is no real title available?)
- Detecting clusters in multivariate response regression
- Selective linearization for multi-block statistical learning
- Learning with Structured Sparsity
- Proximal methods for the latent group lasso penalty
- An efficient Hessian based algorithm for solving large-scale sparse group Lasso problems
- Regularizers for structured sparsity
- Structured variable selection with sparsity-inducing norms
- scientific article; zbMATH DE number 6142618 (Why is no real title available?)
- On the strong convergence of forward-backward splitting in reconstructing jointly sparse signals
- Efficient inexact proximal gradient algorithms for structured sparsity-inducing norm
- Smoothing proximal gradient method for general structured sparse regression
- Mirror Prox algorithm for multi-term composite minimization and semi-separable problems
- Linearized alternating direction method of multipliers for sparse group and fused Lasso models
- Structured sparsity through convex optimization
- Learning with optimal interpolation norms
- scientific article; zbMATH DE number 6378171 (Why is no real title available?)
This page was built for publication: Structured sparsity via alternating direction methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5405166)
