报告题目：Monte Carlo Methods for Statistical Analysis of Tables
报 告 人：陈玉国博士
（美国伊利诺伊大学厄巴纳-尚佩恩分校统计系）
报告时间：2009年12月30日下午4:15
报告地点：管理科研楼1018会议室
举办单位：管理学院统计与金融系、统计研究所
报告内容摘要
In this paper, the authors describe an efficient sequential Monte Carlo method for sampling zero-one tables and multiway contingency tables with given constraints, which can be used to approximate exact conditional inference on such tables. An essential feature of their method is that it samples table entries sequentially according to an appropriate proposal distribution. Computational commutative algebra is used to provide conditions that guarantee that their method has certain good properties. They apply their method to a range of examples from social and medical sciences to demonstrate its efficiency in real problems.
|