A Monte Carlo approach to quantifying model error in Bayesian parameter estimation

被引：2

作者：

White, Staci A. ^{[1
]}

Herbei, Radu ^{[1
]}

机构：

[1] Ohio State Univ, Dept Stat, Columbus, OH 43210 USA

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 2015年 / 83卷

基金：

美国国家科学基金会;

关键词：

phi-divergence; Kullback-Leibler; Hellinger distance; Model error; KULLBACK-LEIBLER; CONVERGENCE; CALIBRATION; INFORMATION; INFERENCE;

D O I：

10.1016/j.csda.2014.10.008

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

Quantifying the discrepancy between two distributions is considered, using the concept of phi-divergence. The motivation is a Bayesian inference scenario where one is interested in comparing different posterior distributions. Strongly consistent estimators for the phi-divergence between two posterior distributions are developed. The proposed estimators alleviate known computational difficulties with estimating normalizing constants. This approach can be used to study the impact that using an approximate likelihood has on the resulting posterior distribution and also to compare the effectiveness of different model approximations. The methodology is applied to two first-order emulator models and an oceanographic application where evaluation of the likelihood function involves the solution to a partial differential equation. (C) 2014 Elsevier B.V. All rights reserved.

引用

页码：168 / 181

页数：14

共 46 条

[1] Wavelet decomposition approaches to statistical inverse problems [J].