Empirical likelihood based Bayesian variable selection

Empirical likelihood is a popular nonparametric statistical tool that does not require any distributional assumptions. The possibility of conducting variable selection via Bayesian empirical likelihood is studied both theoretically and empirically. Theoretically, it is shown that when the prior distribution satisfies certain mild conditions, the corresponding Bayesian empirical likelihood estimators are posteriorly consistent and variable selection consistent. As special cases, the prior of Bayesian empirical likelihood LASSO and SCAD satisfy such conditions and thus can identify the non-zero elements of the parameters with probability approaching 1. In addition, it is easy to verify that those conditions are met for other widely used priors such as ridge, elastic net and adaptive LASSO. Empirical likelihood depends on a parameter that needs to be obtained by numerically solving a non-linear equation. Thus, there exists no conjugate prior for the posterior distribution, which causes the slow convergence of the MCMC sampling algorithm in some cases. To solve this problem, an approximation distribution is used as the proposal to enhance the acceptance rate and, therefore, facilitate faster computation. The computational results demonstrate quick convergence for the examples used in the paper. Both simulations and real data analyses are performed to illustrate the advantages of the proposed methods.