U.S.-Chile Program: Bayesian Simulation and Partially Exchangeable Binary Sequences

Mueller0104496这笔美国和智利的赠款将允许杜克大学的Peter Mueller博士前往智利圣地亚哥，与智利Pontifica University Catolica de智利的Fernando Quintana博士合作。他们的工作涉及使用贝叶斯统计来研究二进制序列形式的数据。具体地说，合作者将研究推理、模型选择和考虑二进制序列时的设计选择问题。这些序列经常出现在处理具有二元可观测的纵向模型时，这在生物医学应用中经常遇到。建模和仿真的主要责任将由智利调查人员负责；与美国调查人员一起负责制定和实施最佳设计方案。

Mueller0104496This US-Chile grant will allow Dr. Peter Mueller of Duke University to travel to Santiago, Chile to work with Dr. Fernando Quintana at Pontifica Universidad Catolica de Chile. Their work concerns the use of Bayesian statistics to study data in the form of binary sequences. Specifically, the collaborators will study issues of inference, model selection, and design choice when binary sequences are considered. These sequences often arise when dealing with longitudinal models with binary observables, often encountered in biomedical applications. Primary responsibility for modeling and simulation will be with the Chilean investigator; responsibility for the development and implementation of the optimal design schemes with the U.S. investigator.