Transductive versions of the LASSO and the Dantzig Selector

TL;DR

This paper introduces transductive variants of the LASSO and Dantzig Selector for high-dimensional linear regression, providing theoretical guarantees and efficient computation, especially for predicting new data points.

Contribution

It proposes generalized, transductive versions of LASSO and Dantzig Selector, extending their applicability and theoretical analysis to new data points.

Findings

The estimators include the classical methods as special cases.

Theoretical guarantees are established under relaxed assumptions.

Efficient algorithms with performance comparable to standard LASSO.

Abstract

We consider the linear regression problem, where the number $p$ of covariates is possibly larger than the number $n$ of observations $(x_{i},y_{i})_{i\leq i \leq n}$, 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 $(x_{i})_{n+1\leq i \leq m}$. 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.