Latent uniform samplers on multivariate binary spaces

Latent uniform samplers on multivariate binary spaces
复制标题

多元二元空间上的潜在均匀采样器

DOI:
10.1007/s11222-023-10276-6
复制
发表时间:
2023
影响因子:
2.2
通讯作者:
Walker, Stephen G.
Walker, Stephen G.
中科院分区:
数学2区
文献类型:
--
作者:
Li, Yanxin;Linero, Antonio;Walker, Stephen G.

文献摘要

参考文献

相似文献

我们考虑从一个概率分布或一个等价的高维二进制空间中采样。许多重要的应用依赖于从这样的分布中采样,包括贝叶斯变量选择问题和拟合贝叶斯回归树。当维数很大时,由于存在可能的状态,直接采样是禁止的。对这种分布进行采样的一种方法是使用Metropolis-Hastings算法,这可能需要选择一个合适的建议机制,默认选择是单组件交换机建议移动。当存在多个模式时,这是有问题的。在本文中,我们提出了一个潜在的变量均匀采样算法,如潜在的切片采样器,它允许大的移动和建议路径,提供不可忽略的概率模式之间的移动,即使当这些模式之间的状态的概率是低的。提出了一些插图,主要集中在演示当前的通用采样器的优势。
We consider sampling from a probability distribution on, or an equivalent high-dimensional binary space. A number of important applications rely on sampling from such distributions, including Bayesian variable selection problems and fitting Bayesian regression trees. Direct sampling is prohibitive when the dimension is large due to the fact that there arepossible states. One approach to sampling such distributions is to use a Metropolis–Hastings algorithm, which can require choosing a decent proposal mechanism, with a default choice being the single-component switch proposal move. This is problematic when multiple modes exist. In this paper, we propose a latent variable uniform sampling algorithm, such as a latent slice sampler, which allows for large moves and proposal paths which give non-negligible probabilities for moving between modes, even when the probabilities of states between these modes is low. A number of illustrations are presented, focusing primarily on demonstrating the advantages over current generic samplers.
DOI: 10.1214/19-aos1916
发表时间: 2018-12
期刊: The Annals of Statistics
影响因子: --
作者:
E. Pompe;C. Holmes;K. Latuszy'nski
通讯作者: E. Pompe;C. Holmes;K. Latuszy'nski
带追索权的离散随机优化的增强仿真方法
DOI: --
发表时间: 2020
影响因子: 4.8
作者:
Tahir Ekin;S. Walker;P. Damien
通讯作者: P. Damien
概率:都市-黑斯廷斯算法的替代方案
DOI: --
发表时间: 2014
期刊:
影响因子: --
作者:
S. Walker;S. Walker
通讯作者: S. Walker
DOI: 10.1080/01621459.2022.2037431
发表时间: 2022-03-17
影响因子: 3.7
作者:
Li, Yinpu;Linero, Antonio R.;Murray, Jared
通讯作者: Murray, Jared
使用软贝叶斯加性回归树(SBART)对聚类区间删失生存数据进行半参数分析
DOI: 10.1111/biom.13478
发表时间: 2020
期刊: Biometrics
影响因子: 1.9
作者:
Piyali Basak;A. Linero;D. Sinha;S. Lipsitz
通讯作者: S. Lipsitz