The Nested Dirichlet Process
The Nested Dirichlet Process
复制标题
DOI：
10.1198/016214508000000553
复制
发表时间：
2008-09-01
影响因子：
3.7
通讯作者：
Gelfand, Alan E.
中科院分区：
文献类型：
--
作者：
Rodriguez, Abel;Dunson, David B.;Gelfand, Alan E.
在多中心研究中，不同中心的研究对象可能有不同的结果分布。本文的动机是这些分布的非参数建模问题，借用跨中心的信息，同时也允许中心聚类，从Dricichlet过程（DP）的棍子断裂表示开始。我们用从DP中得到的随机概率度量来代替随机原子。这导致了一个嵌套的DP先验，它可以放在不同中心的分布集合上，这些中心从相同的DP组件中绘制而来，自动聚集在一起。讨论了理论性质，提出了一种有效的马尔可夫链蒙特卡罗算法。通过模拟研究和美国医院护理质量的应用说明了这些方法。
In multicenter studies subjects in different centers may have different outcome distribution. This article is motivated by the problem of nonparametric modeling of these distributions, borrowing information across centers while also allowing centers to be clustered, Starting with a stick-breaking representation of the Dricichlet process (DP). we replace that random atoms with random probability measures drawn from a DP. This results in a nested DP prior, which can be placed on the collection of distributions for the different centers with centers drawn from the same DP component authomatically clustered together. Theorectical properties are discussed and an efficient Markov chain Monte Carlo algorithm is developed for computation. The methods are illustrated using a simulation study and an application to quality of care in U.S hospitals.