Improved LARS Algorithm for Adaptive LASSO in the Linear Regression Model

Abstract

The adaptive LASSO method has been employed for reliable variable selection as an alternative to LASSO in linear regression models. This paper introduces an adjusted LARS algorithm that integrates adaptive LASSO with several biased estimators, including the Almost Unbiased Ridge Estimator (AURE), Liu Estimator (LE), Almost Unbiased Liu Estimator(AULE), Principal Component Regression Estimator (PCRE), r-k class estimator, and r-d class estimator. The effectiveness of the proposed algorithm is evaluated through Monte Carlo simulation and empirical examples.