Bayesian High-dimensional Semi-parametric Inference beyond sub-Gaussian Errors

2020/08/30 by Lee, Kyoungjae, Chae, Minwoo, Lin, Lizhen
#FOS: Mathematics #Statistics Theory (math.ST)

paper · doi:10.48550/arxiv.2008.13174

Abstract

We consider a sparse linear regression model with unknown symmetric error under the high-dimensional setting. The true error distribution is assumed to belong to the locally β-Hölder class with an exponentially decreasing tail, which does not need to be sub-Gaussian. We obtain posterior convergence rates of the regression coefficient and the error density, which are nearly optimal and adaptive to the unknown sparsity level. Furthermore, we derive the semi-parametric Bernstein-von Mises (BvM) theorem to characterize asymptotic shape of the marginal posterior for regression coefficients. Under the sub-Gaussianity assumption on the true score function, strong model selection consistency for regression coefficients are also obtained, which eventually asserts the frequentist's validity of credible sets.

Related

• Laplace Approximation in High-dimensional Bayesian Regression
• Bayesian High-dimensional Linear Regression with Sparse Projection-posterior
• Posterior Asymptotic Normality for an Individual Coordinate in High-dimensional Linear Regression
• Semiparametric posterior limits
• Inference in high-dimensional regression models without the exact or L^p sparsity