MALA-within-Gibbs Samplers for High-Dimensional Distributions with Sparse Conditional Structure
MALA-within-Gibbs Samplers for High-Dimensional Distributions with Sparse Conditional Structure
复制标题
用于具有稀疏条件结构的高维分布的 MALA-in-Gibbs 采样器
DOI:
10.1137/19m1284014
复制
发表时间:
2020
影响因子:
3.1
通讯作者:
Marzouk, Y. M.
中科院分区:
文献类型:
--
作者:
Tong, X. T.;Morzfeld, M.;Marzouk, Y. M.
Markov chain Monte Carlo (MCMC) samplers are numerical methods for drawing samples from a given target probability distribution. We discuss one particular MCMC sampler, the MALA-within-Gibbs sampler, from the theoretical and practical perspectives. We first show that the acceptance ratio and step size of this sampler are independent of the overall problem dimension when (i) the target distribution has sparse conditional structure, and (ii) this structure is reflected in the partial updating strategy of MALA-within-Gibbs. If, in addition, the target density is blockwise log-concave, then the sampler's convergence rate is independent of dimension. From a practical perspective, we expect that MALA-within-Gibbs is useful for solving high-dimensional Bayesian inference problems where the posterior exhibits sparse conditional structure at least approximately. In this context, a partitioning of the state that correctly reflects the sparse conditional structure must be found, and we illustrate this process in two numerical examples. We also discuss trade-offs between the block size used for partial updating and computational requirements that may increase with the number of blocks.
登录
查看更多内容
DOI:
10.1007/978-1-4899-0319-8_6
发表时间:
1997
期刊:
--
影响因子:
--
作者:
A. Sokal
通讯作者:
A. Sokal
影响因子:
2.4
作者:
A. Chorin;O. Hald
通讯作者:
O. Hald
DOI:
10.1111/j.1467-9868.2005.00500.x
发表时间:
2005
期刊:
Journal of the Royal Statistical Society: Series B (Statistical Methodology)
影响因子:
--
作者:
O. F. Christensen;G. Roberts;J. Rosenthal
通讯作者:
J. Rosenthal
DOI:
--
发表时间:
1991
期刊:
影响因子:
--
作者:
R. Edwards;J. Goodman;A. Sokal
通讯作者:
A. Sokal
DOI:
10.1029/2008gm000763
发表时间:
2013
期刊:
SIAM J. Appl. Dyn. Syst.
影响因子:
--
作者:
C. Jamet;H. Loisel
通讯作者:
H. Loisel