105
49

A Bernstein-Von Mises Theorem for discrete probability distributions

Abstract

We investigate the asymptotic normality of the posterior distribution in the discrete setting, when model dimension increases with sample size. We consider a probability mass function θ0\theta_0 on \mathbbmN{0}\mathbbm{N}\setminus \{0\} and a sequence of truncation levels (kn)n(k_n)_n satisfying kn3ninfiknθ0(i).k_n^3\leq n\inf_{i\leq k_n}\theta_0(i). Let θ^\hat{\theta} denote the maximum likelihood estimate of (θ0(i))ikn(\theta_0(i))_{i\leq k_n} and let Δn(θ0)\Delta_n(\theta_0) denote the knk_n-dimensional vector which ii-th coordinate is defined by \sqrt{n} (\hat{\theta}_n(i)-\theta_0(i)) for 1ikn.1\leq i\leq k_n. We check that under mild conditions on θ0\theta_0 and on the sequence of prior probabilities on the knk_n-dimensional simplices, after centering and rescaling, the variation distance between the posterior distribution recentered around θ^n\hat{\theta}_n and rescaled by n\sqrt{n} and the knk_n-dimensional Gaussian distribution N(Δn(θ0),I1(θ0))\mathcal{N}(\Delta_n(\theta_0),I^{-1}(\theta_0)) converges in probability to 0.0. This theorem can be used to prove the asymptotic normality of Bayesian estimators of Shannon and R\'{e}nyi entropies. The proofs are based on concentration inequalities for centered and non-centered Chi-square (Pearson) statistics. The latter allow to establish posterior concentration rates with respect to Fisher distance rather than with respect to the Hellinger distance as it is commonplace in non-parametric Bayesian statistics.

View on arXiv
Comments on this paper