Shrinkage estimators' properties in regression models using L_1 penalized norm

نویسندگان
Ghoreishi, Seyed Kamran ID 1

1 Department of Statistics, University of Qom, Qom, Iran

doi
10.22091/maa.2023.9446.1003
چکیده

In this paper, we first introduce a two-level hierarchical model with a linear structure.  We use the moment, maximum likelihood, and SURE estimators to obtain the regression coefficients shrinkage estimators. Since, regression models have vast applications in high-dimensional datasets, using sparsity assumption, we discuss the asymptotic properties of the regression estimators under L_2 error norm and L_1 penalty norm. 

کلیدواژه‌ها
Hierarchical models, Shrinkage estimators, Regression models, High-dimensional datasets