A new Bayesian elastic net for tobit regression

Abstract

Abstract In this paper, we propose a new Bayesian elastic net (EN) approach for variable selection and coefficient estimation in tobit regression. Specifically, we present a new hierarchical formulation of the Bayesian EN by utilizing the scale mixture of truncated normal distribution (with exponential mixing distributions) of the Laplace density part. The proposed method is an alternative method to Bayesian method of the EN problem. The performance of the proposed model is compared with old model of the Bayesian elastic net using a simulation example. It is shown that the model performs well compared with old elastic net representation.