This paper focuses on the numerical computation of posterior expected quantities of interest, where existing approaches based on ergodic averages are gated by the asymptotic variance of the integrand. To address this challenge, a novel technique is proposed to post-process Markov chain Monte Carlo output, based on Sard's approach to numerical integration and the control functional method. The use of Sard's approach ensures that our control functionals are exact on all polynomials up to a fixed degree in the Bernstein-von-Mises limit, so that the reduced variance estimator approximates the behaviour of a polynomially-exact (e.g. Gaussian) cubature method. The proposed method is shown to combine the robustness of parametric control variates with the flexibility of non-parametric control functionals across a selection of Bayesian inference tasks. All methods used in this paper are available in the R package ZVCV.
翻译:本文侧重于后期预期利益量的计算, 现有基于egodic平均值的办法被原数的无症状差异所束缚。 为了应对这一挑战, 根据Sard的数值整合方法和控制功能方法, 提议了马可夫链蒙特卡洛后处理后产出的新技术。 使用 Sard 方法可以确保我们的控制功能精确到 Bernstein- von-Mises 限值中以固定程度为限的所有多数值, 以便减少的差异估计器能够接近多球形( 如高山) 肿瘤方法的行为。 拟议的方法可以将参数控制变量的稳健性与非参数控制功能的灵活性结合到选择 Bayesian 推断任务中。 本文使用的所有方法都可以在 R 包 ZVCV 中查阅 。