A Permutation Approach for Selecting the Penalty Parameter in Penalized Model Selection

TL;DR

This paper introduces a permutation-based method for selecting the penalty parameter in LASSO, aiming to improve variable selection accuracy across various models, and compares it with existing methods through simulations and real data analysis.

Contribution

It presents a novel permutation approach for penalty selection in LASSO, applicable to diverse structural models, and evaluates its performance against traditional methods.

Findings

Permutation method performs comparably or better than CV and BIC.

The approach is effective in generalized linear models.

Simulation and real data analyses validate its utility.

Abstract

We describe a simple, efficient, permutation based procedure for selecting the penalty parameter in the LASSO. The procedure, which is intended for applications where variable selection is the primary focus, can be applied in a variety of structural settings, including generalized linear models. We briefly discuss connections between permutation selection and existing theory for the LASSO. In addition, we present a simulation study and an analysis of three real data sets in which permutation selection is compared with cross-validation (CV), the Bayesian information criterion (BIC), and a selection method based on recently developed testing procedures for the LASSO.