BriefGPT.xyz
Jul, 2017
使用DoGS改善Gibbs采样器扫描质量
Improving Gibbs Sampler Scan Quality with DoGS
HTML
PDF
Ioannis Mitliagkas, Lester Mackey
TL;DR
使用 Dobrushin-optimized Gibbs samplers (DoGS)为基础，提供量身定做的变量选择顺序，在总变异距离到稳态方面提供了明确的上下限，明显降低了抽样预算，并证实了在联合图像分割和物体识别等方面的实际优越性。
Abstract
The pairwise influence matrix of Dobrushin has long been used as an analytical tool to bound the rate of convergence of
gibbs sampling
. In this work, we use
dobrushin influence
as the basis of a practical tool to
→