New scale mixture for Bayesian adaptive lasso Tobit regression

Abstract EN

Abbas [1] proposed new hierarchical representation of the adaptive Bayesian lasso model as uniform density, mixing with standard exponential distribution based on a transformation of the mixture of uniform density and a particular gamma distribution formulation provided by Mallick & Yi [2] They consider the new proposed hierarchical formulation model and prior distributions, as well as the full Conditional posterior distributions structural under non conditioning on σ2 which makes the uncertainty, of a unimodal full posterior, Conditioning on σ2 is important, because it guarantees a unimodal full posterior Park and Casella [3].

So, we can conclude that [1] proposed new hierarchical representation utilizing a Non- scale mixture distributions, which needs to deal with this problem.

To address this problem we consider new hierarchical representation of the adaptive Bayesian lasso for Tobit model based on scale mixture of Uniform density, mixing with standard exponential distribution.