New penalized Bayesian adaptive lasso binary regression

Abstract

The scale mixture of normal mixing with Rayleigh as representation of Laplace prior of β has introduced by Flaih et al[1].We employed this new scale mixture for the adaptive lasso Binary regression. New hierarchical model is considering ,as well the Gibbs sampler algorithm in introduced . We considering the new penalized Bayesian adaptive lasso in Binary regression as variable selection method in case of presenting they high dimensional data . The new proposed model can overcame the multicollinearity problem in predictor variables. We conducting simulation analysis, as well as real data application to show the performance of the proposed method.