Label Switching in Bayesian Mixture Models: Deterministic Relabeling Strategies
Label Switching in Bayesian Mixture Models: Deterministic Relabeling Strategies
复制标题
贝叶斯混合模型中的标签切换：确定性重新标记策略
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
S. Walker
中科院分区：
文献类型：
--
作者：
Carlos E. Rodríguez;S. Walker
标签切换是混合模型贝叶斯分析中的一个众所周知的问题。一方面，它使推理变得复杂，另一方面，它被认为是证明马尔可夫链蒙特卡罗(MCMC)收敛的先决条件。因此，已经提出了遍历后验分布的对称副本以及可能的真模式的非标准MCMC算法。为了执行特定于组件的推断，需要应用撤销标签交换和恢复组件解释的方法。如果包括用于设计MCMC策略的潜在分配，并且采样器已经收敛，则分配给每个组件的标签可能随着迭代而改变。然而，被分配在一起的观测值必须保持相似，我们使用这一基本事实来推导出标签交换问题的简单而有效的解决方案。我们在单变量和多变量数据示例上将我们的策略与其他重新标记算法进行了比较，并展示了对替代策略的改进。这篇文章的补充材料可以在网上找到。
Label switching is a well-known problem in the Bayesian analysis of mixture models. On the one hand, it complicates inference, and on the other hand, it has been perceived as a prerequisite to justify Markov chain Monte Carlo (MCMC) convergence. As a result, nonstandard MCMC algorithms that traverse the symmetric copies of the posterior distribution, and possibly genuine modes, have been proposed. To perform component-specific inference, methods to undo the label switching and to recover the interpretation of the components need to be applied. If latent allocations for the design of the MCMC strategy are included, and the sampler has converged, then labels assigned to each component may change from iteration to iteration. However, observations being allocated together must remain similar, and we use this fundamental fact to derive an easy and efficient solution to the label switching problem. We compare our strategy with other relabeling algorithms on univariate and multivariate data examples and demonstrate improvements over alternative strategies. Supplementary materials for this article are available online.
DOI：
--
发表时间：
1997
期刊：
Journal of the royal statistical society series b-methodological
影响因子：
--
作者：
S. Richardson;P. Green;Christian P. Robert;M. Aitkin;David R. Cox;Matthew Stephens;A. Polymenis;W. Gilks;A. Nobile;M. Hodgson;Anthony O'Hagan;N. Longford;A. Dawid;Anthony C. Atkinson;J. Bernardo;J. Besag;Stephen Brooks;S. Byers;A. Raftery;G. Celeux;R. Cheng;W. B. Liu;Yung-Hsin Chien;Edward I. George;N. Cressie;H.-C. Huang;M. Gruet;S. C. Heath;C. Jennison;Andrew B. Lawson;Allan Clark;Geoffrey J. McLachlan;D. Peel;K. Mengersen;A. George;Anne Philippe;Kathryn Roeder;Larry Wasserman;Peter Schlattmann;D. Böhning;D. M. Titterington;H. Tong;M. West
通讯作者：
S. Richardson;P. Green;Christian P. Robert;M. Aitkin;David R. Cox;Matthew Stephens;A. Polymenis;W. Gilks;A. Nobile;M. Hodgson;Anthony O'Hagan;N. Longford;A. Dawid;Anthony C. Atkinson;J. Bernardo;J. Besag;Stephen Brooks;S. Byers;A. Raftery;G. Celeux;R. Cheng;W. B. Liu;Yung-Hsin Chien;Edward I. George;N. Cressie;H.-C. Huang;M. Gruet;S. C. Heath;C. Jennison;Andrew B. Lawson;Allan Clark;Geoffrey J. McLachlan;D. Peel;K. Mengersen;A. George;Anne Philippe;Kathryn Roeder;Larry Wasserman;Peter Schlattmann;D. Böhning;D. M. Titterington;H. Tong;M. West