文章基本信息

标题：A Hierarchical Bayes Approach to Estimation and Prediction for Time Series of Counts
本地全文：下载
作者：P. J. FARRELL ; B. MACGIBBON ; T. J. TOMBERLIN 等
期刊名称：Brazilian Journal of Probability and Statistics
印刷版ISSN：0103-0752
出版年度：2007
卷号：21
期号：2
页码：187-202
出版社：Brazilian Statistical Association
摘要：In this paper, we are concerned with the statistical method-ology of epidemiological surveillance; that is, the ongoing procedure of an-alyzing and interpreting public health data of infectious disease incidence.In particular, we propose a hierarchical Bayes approach for the estimationof generalized linear mixed models for time series count data, and their usein the prediction of counts for future time periods. The estimators are ob-tained by Gibbs sampling and their performance is compared to those of othermethods on the polio data originally analysed by Zeger (1988), which consistof the monthly number of U.S. polio cases between 1970 and 1983. Theirproperties are also investigated via simulation. Our aim is to illustrate howeasily the hierarchical Bayes methodology lends itself to mo del checking andmodel comparisons. The proposed methodology, in particular, hierarchicalBayes prediction, is applied to a series of Campylobacter infection cases inthe Montreal-Center region
关键词：Adaptive rejection Metropolis sampling; epidemiological;surveillance; Gibbs sampling; infectious diseases; longitudinal data; Poisson;counts