Mathematical Sciences: CBMS Reg Conf -Semiparametic MixturesModels: Theory and Methods with Applications in Empirical Bayes, Measurements Error and Random Effects Models - May 93

点击翻译按钮获取中文摘要

Recently there has been substantial research on mixture models based on a nonparametric mixing distribution, also called semiparametric mixture models. Applications have ranged from clinical trials, astronomy and marine biology. Within statistics, research from semiparametric mixture models extends to empirical Bayes' estimation, measurement error models and random effects models, as well as other fields. Insights into these questions and others have been greatly aided by work done by the principal speaker, Bruce Lindsay. Professor Lindsay 's research on the geometry of mixture likelihoods has provided a unified approach for interpreting, solving and formulating models of mixed distributions. His intuitive approach and geometric insights make him an ideal lecturer for an NSF/CBMS conference on the subject. This project will support an NSF-CBMS Regional Research Conference in the Mathematical Sciences on Semiparametric Mixture Methods: Theory and Methods - With Applications in Bayes, Measurment Error, and Random Effects Models to be held May 24 - 28, 1993, on the campus of the University of South Carolina. Professor Bruce Lindsay of the Pennsylvania State University will be the principal lecturer. To stimulate interest and activity in mathematical research, the National Science Foundation each year supports a number of NSF-CBMS Regional Research Conferences in the Mathematical Sciences. Each five-day conference features a distinguished lecturer who delivers ten lectures on a topic on important current research in one sharply focused area of the mathematical sciences. The lecturer subsequently prepares an expository monograph based upon these lectures, which is normally published by the American Mathematical Society or the Society for Industrial and Applied Mathematics, or jointly by the American Statistical Association and the Institute of Mathematical Statistics. Certain features differentiate these conferences from typical research conferences. These are: (1) Focus on a single important and timely area of research by a leading practitioner, (2) Continued effect and local stimulation through regional emphasis, (3) Panel review for quality, breadth, and timeliness, and (4) Published monographs for a wider audience.