The Polya Tree Sampler: Towards Efficient and Automatic Independent Metropolis-Hastings Proposals.
The Polya Tree Sampler: Towards Efficient and Automatic Independent Metropolis-Hastings Proposals.
复制标题
polya树采样器：朝着高效且自动独立的大都市 - 危机提案。
DOI：
10.1198/jcgs.2010.09115
复制
发表时间：
2011-03-01
期刊：
影响因子：
--
通讯作者：
Jara A
中科院分区：
文献类型：
--
作者：
Hanson TE;Monteiro JV;Jara A
我们提出了一种简单，高效，计算成本低的采样方法，用于探索非归一化的多元密度上，如后验密度，称为波利亚树采样器。该算法构造一个独立的建议的基础上的目标密度的近似。近似是从一组（初始）支持点（作为近似参数的数据）和有限多变量Polya树的预测密度构建的。在初始“预热”阶段，支持点被迭代地重新定位到目标分布下的较高支持的区域，以最小化目标分布和Polya树预测分布之间的距离。在“抽样”阶段，从有限波利亚树的最终近似混合物的样本被用作候选人，这是接受一个标准的大都会黑斯廷斯接受概率。文中给出了几个算例，并与Metropolis-within-Gibbs算法和延迟拒绝自适应大都会算法进行了比较。
We present a simple, efficient, and computationally cheap sampling method for exploring an un-normalized multivariate density on ℝd, such as a posterior density, called the Polya tree sampler. The algorithm constructs an independent proposal based on an approximation of the target density. The approximation is built from a set of (initial) support points – data that act as parameters for the approximation – and the predictive density of a finite multivariate Polya tree. In an initial “warming-up” phase, the support points are iteratively relocated to regions of higher support under the target distribution to minimize the distance between the target distribution and the Polya tree predictive distribution. In the “sampling” phase, samples from the final approximating mixture of finite Polya trees are used as candidates which are accepted with a standard Metropolis-Hastings acceptance probability. Several illustrations are presented, including comparisons of the proposed approach to Metropolis-within-Gibbs and delayed rejection adaptive Metropolis algorithm.
登录
查看更多内容
DOI：
10.1137/0908055
发表时间：
1987-07-01
期刊：
SIAM JOURNAL ON SCIENTIFIC AND STATISTICAL COMPUTING
影响因子：
--
作者：
ANDERSON, TW;OLKIN, I;UNDERHILL, LG
通讯作者：
UNDERHILL, LG
影响因子：
2.4
作者：
Giordani, Paolo;Kohn, Robert
通讯作者：
Kohn, Robert
DOI：
10.1080/01621459.1995.10476590
发表时间：
1995-09-01
影响因子：
3.7
作者：
GEYER, CJ;THOMPSON, EA
通讯作者：
THOMPSON, EA
影响因子：
1
作者：
Hanson, Timothy E.;Branscum, Adam J.;Gardner, Ian A.
通讯作者：
Gardner, Ian A.
影响因子：
2.2
作者：
Cappe, Olivier;Douc, Randal;Robert, Christian P.
通讯作者：
Robert, Christian P.