Sparsity with sign-coherent groups of variables via the cooperative-Lasso

DOI10.1214/11-AOAS520zbMATH Open1243.62101arXiv1103.2697MaRDI QIDQ439175FDOQ439175

Authors: Yves Grandvalet, Camille Charbonnier, Julien Chiquet 

Publication date: 1 August 2012

Published in: The Annals of Applied Statistics (Search for Journal in Brave)

Abstract: We consider the problems of estimation and selection of parameters endowed with a known group structure, when the groups are assumed to be sign-coherent, that is, gathering either nonnegative, nonpositive or null parameters. To tackle this problem, we propose the cooperative-Lasso penalty. We derive the optimality conditions defining the cooperative-Lasso estimate for generalized linear models, and propose an efficient active set algorithm suited to high-dimensional problems. We study the asymptotic consistency of the estimator in the linear regression setup and derive its irrepresentable conditions, which are milder than the ones of the group-Lasso regarding the matching of groups with the sparsity pattern of the true parameters. We also address the problem of model selection in linear regression by deriving an approximation of the degrees of freedom of the cooperative-Lasso estimator. Simulations comparing the proposed estimator to the group and sparse group-Lasso comply with our theoretical results, showing consistent improvements in support recovery for sign-coherent groups. We finally propose two examples illustrating the wide applicability of the cooperative-Lasso: first to the processing of ordinal variables, where the penalty acts as a monotonicity prior; second to the processing of genomic data, where the set of differentially expressed probes is enriched by incorporating all the probes of the microarray that are related to the corresponding genes.

Full work available at URL: https://arxiv.org/abs/1103.2697

Recommendations

• The Group Lasso for Logistic Regression
• Consistent group selection in high-dimensional linear regression
• Adaptive group Lasso for high-dimensional generalized linear models
• Coordinate descent algorithms for lasso penalized regression
• On the asymptotic properties of the group lasso estimator for linear models

zbMATH Keywords

penalizationcontinuous variablesmicroarray analysisordinal variablesgrouped variables

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Linear regression; mixed models (62J05) Generalized linear models (logistic models) (62J12) Graphical methods in statistics (62A09)

Cites Work

• Inferring multiple graphical structures
• A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
• Heuristics of instability and stabilization in model selection
• Title not available (Why is that?)
• Asymptotic Statistics
• Title not available (Why is that?)
• Estimation of the mean of a multivariate normal distribution
• Regularization and Variable Selection Via the Elastic Net
• Sparse modeling of categorial explanatory variables
• Model Selection and Estimation in Regression with Grouped Variables
• Asymptotics for Lasso-type estimators.
• The Group Lasso for Logistic Regression
• Least angle and \(\ell _{1}\) penalized regression: a review
• The composite absolute penalties family for grouped and hierarchical variable selection
• Better Subset Regression Using the Nonnegative Garrote
• On the Non-Negative Garrotte Estimator
• An active set algorithm to estimate parameters in generalized linear models with ordered predictors
• On the ``degrees of freedom of the lasso
• Consistency of the group Lasso and multiple kernel learning
• The benefit of group sparsity
• The Estimation of Prediction Error
• On the asymptotic properties of the group lasso estimator for linear models
• Averaged gene expressions for regression
• Using cluster computers in bioinformatics research
• Defining a robust biological prior from pathway analysis to drive network inference

Cited In (7)

• ``Grouping strategies and thresholding for high dimensional linear models: rejoinder
• Grouping strategies and thresholding for high dimensional linear models
• Additive monotone regression in high and lower dimensions
• Monotone splines Lasso
• Regularization and model selection for quantile varying coefficient model with categorical effect modifiers
• Partially linear monotone methods with automatic variable selection and monotonicity direction discovery
• A Two-Part Framework for Estimating Individualized Treatment Rules From Semicontinuous Outcomes

Uses Software

• UCI-ml

This page was built for publication: Sparsity with sign-coherent groups of variables via the cooperative-Lasso

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q439175)