Transductive versions of the Lasso and the Dantzig selector

DOI10.1016/J.JSPI.2012.03.020zbMATH Open1428.62312arXiv0906.0652OpenAlexW2964240626MaRDI QIDQ447611FDOQ447611

Authors: Pierre Alquier, Mohamed Hebiri 

Publication date: 4 September 2012

Published in: Journal of Statistical Planning and Inference (Search for Journal in Brave)

Abstract: We consider the linear regression problem, where the number $p$ of covariates is possibly larger than the number $n$ of observations $(x_i,y_i{)}_{ileqileqn}$, under sparsity assumptions. On the one hand, several methods have been successfully proposed to perform this task, for example the LASSO or the Dantzig Selector. On the other hand, consider new values $(x_i{)}_{n+1leqileqm}$. If one wants to estimate the corresponding $y_i$'s, one should think of a specific estimator devoted to this task, referred by Vapnik as a "transductive" estimator. This estimator may differ from an estimator designed to the more general task "estimate on the whole domain". In this work, we propose a generalized version both of the LASSO and the Dantzig Selector, based on the geometrical remarks about the LASSO in pr'evious works. The "usual" LASSO and Dantzig Selector, as well as new estimators interpreted as transductive versions of the LASSO, appear as special cases. These estimators are interesting at least from a theoretical point of view: we can give theoretical guarantees for these estimators under hypotheses that are relaxed versions of the hypotheses required in the papers about the "usual" LASSO. These estimators can also be efficiently computed, with results comparable to the ones of the LASSO.

Full work available at URL: https://arxiv.org/abs/0906.0652

Recommendations

• On the prediction loss of the Lasso in the partially labeled setting
• Simultaneous analysis of Lasso and Dantzig selector
• The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
• scientific article; zbMATH DE number 2243369
• Persistene in high-dimensional linear predictor-selection and the virtue of overparametrization

zbMATH Keywords

Lassovariable selectionhigh-dimensional datasparsityhigh-dimensional regression estimationtransduction

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Linear regression; mixed models (62J05) Ridge regression; shrinkage estimators (Lasso) (62J07)

Cites Work

• High-dimensional additive modeling
• Estimating the dimension of a model
• The Adaptive Lasso and Its Oracle Properties
• Least angle regression. (With discussion)
• Pathwise coordinate optimization
• Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
• Title not available (Why is that?)
• Title not available (Why is that?)
• Lasso-type recovery of sparse representations for high-dimensional data
• On the conditions used to prove oracle results for the Lasso
• Simultaneous analysis of Lasso and Dantzig selector
• High-dimensional generalized linear models and the lasso
• The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
• High-dimensional graphs and variable selection with the Lasso
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Text classification from labeled and unlabeled documents using EM
• DASSO: Connections Between the Dantzig Selector and Lasso
• Title not available (Why is that?)
• On the Non-Negative Garrotte Estimator
• Aggregation for Gaussian regression
• Consistency of the group Lasso and multiple kernel learning
• The Dantzig selector and sparsity oracle inequalities
• Some theoretical results on the grouped variables Lasso
• Sup-norm convergence rate and sign concentration property of Lasso and Dantzig estimators
• Title not available (Why is that?)
• On transductive support vector machines
• Smoothing \(\ell_1\)-penalized estimators for high-dimensional time-course data
• Sparse recovery in convex hulls via entropy penalization
• Aggregation by Exponential Weighting and Sharp Oracle Inequalities
• Lasso, iterative feature selection and the correlation selector: oracle inequalities and numerical performances
• Generalization of constraints for high dimensional regression problems

Cited In (3)

• On the prediction loss of the Lasso in the partially labeled setting
• DASSO: Connections Between the Dantzig Selector and Lasso
• Optimal and Safe Estimation for High-Dimensional Semi-Supervised Learning

Uses Software

• hgam

This page was built for publication: Transductive versions of the Lasso and the Dantzig selector

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q447611)