Fifth International Workshop on Objective Bayesian Methodology

第五届客观贝叶斯方法论国际研讨会将于2005年6月4日至8日在密苏里州的布兰森举行。客观贝叶斯方法论在很大程度上是面向可自动使用的先验分布的发展，即除了选择描述数据的特定概率模型外，不需要主观输入。这一发展发生在三个截然不同的统计领域：参数估计、模型选择和预测。随着贝叶斯分析在非专业人士中的应用迅速增长，目标贝叶斯方法在今天变得越来越重要。大多数非专家寻求自动或客观的贝叶斯程序。本课程将侧重于预测和实际应用，如时空模型、多重比较和拟合优度。贝叶斯方法通常严重依赖于先验分布。尽管先验分布是贝叶斯范式中的一个重要组成部分，但众所周知，即使在缺乏任何相关的先验信息的情况下，贝叶斯方法也可以通过使用适当的客观(非信息)贝叶斯分析来对许多具有挑战性的问题做出可靠的推断。主要目标是促进交流客观贝叶斯方法论的最新研究进展，为新的研究人员提供机会，并建立新的合作，将努力引导到悬而未决的问题上，并为调查开辟新的方向。这次会议的一个重要结果将是进一步明确客观优先方法论作为统计研究的既定领域。统计学家传统上接受两种主要方法之一，即经典或“频率”方法和贝叶斯方法。近年来，计算技术的进步和统计理论的突破使得贝叶斯方法越来越有吸引力，特别是对于气象学、流行病学、自然资源等领域的多变量难题。随着对贝叶斯统计学的日益重视，弥合这两个统计思想流派之间的差异变得越来越重要。客观贝叶斯致力于解决这些差异。计划中的讲习班将汇集80-100名与会者，其中包括来自世界各地的客观贝叶斯统计领域的领导人，以及青年研究人员和学生，在一个特别有利于合作和思想交流的环境中。

The Fifth International Workshop on Objective Bayesian Methodology willbe held in Branson, Missouri, June 4-8, 2005.Objective Bayesian methodology is, for the most part, oriented towardsthe development of prior distributions that can be used automatically,i.e. that do not require subjective input other than the specificprobabilistic model chosen to describe the data. There are three quitedistinct statistical domains in which this development has taken place:parametric estimation, model selection, and prediction,Objective Bayesian methodology is of increasing importance today sinceapplication of Bayesian analysis is rapidly growing among nonspecialists,most of whom seek automatic or objective Bayesian procedures.This workshop will emphasize prediction, practical applicationssuch as spatial-temporal models, multiple comparisons and goodness-of-fit.Bayesian methods usually are heavily depend on prior distributions.Although a prior distribution is an important component in the Bayesianparadigm, it is known that even in the absence of any relevant priorinformation, the Bayes method can produce reliable inference to manychallenging problems by employing appropriate objective (noninformative)Bayesian analysis. The principal objectives are to facilitate theexchange of recent research developments in objective Bayesianmethodology, to provide opportunities for new researchers, and toestablish new collaborations that will channel efforts into pendingproblems and open new directions for investigation. An importantconsequence of this meeting will be to further crystallize objective priormethodology as an established area for statistical research.Statisticians have traditionally embraced one of two main approaches,classical or "frequentist" methods and Bayesian methods. In recent years,advances in computing technology and breakthroughs in statistical theoryhave made Beyesian methods increasingly attractive, especially for difficult problems with many variables in fields such as meteorology, epidemiology, natural resources, etc. With this increased emphasis on Bayesian statistics, it has become increasingly important to bridge the differences between the two schools of statistical thinking. Objective Bayesians are dedicated to resolving these differences. The planned workshop will bring together 80-100 participants including leaders in the field of objective Bayesian statistics from around the world together with young researchers and students in a setting especially conducive to cooperation and exchange of ideas.