Asymptotic Analysis of Adaptive Lasso, Transfer Lasso, and a Novel Integrated Method

The Adaptive Lasso and Transfer Lasso employ initial estimators in different ways, leading to distinct asymptotic properties. The Adaptive Lasso achieves both √n-consistency and consistent variable selection, while the Transfer Lasso can attain faster convergence rates but lacks consistent variable selection. A novel Adaptive Transfer Lasso method is proposed that integrates the strengths of both approaches.