Collaborative Research on Flexible Modeling and Analysis for Information Aggregation

信息聚合灵活建模与分析协同研究

• 批准号: 9022616
• 负责人: Robert Clemen
• 金额: $ 13.37万
• 依托单位: University of Oregon Eugene
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1991
• 资助国家: 美国
• 起止时间: 1991-02-15 至 1994-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9022616&HistoricalAwards=false
• 关键词: Collaborative; Research; Flexible; Modeling; Analysis

In many problems, information is available from many different sources. Aggregating these diverse bits of information is important to good decision-making but also leads to special statistical problems in developing models to characterize the information. Prior research in this areas has relied primarily on the use of historical data as a basis for modelling the information sources. The proposed research develops a Bayesian framework that a decision maker can use to encode his subjective knowledge about the information sources in order to aggregate their information. This framework features a highly flexible environment for modelling the probabilistic nature and interrelationships of the information sources and requires straightforward and intuitive subjective judgements using proven decision-analytic assessment techniques. Analysis of the assessments to produce a posterior distribution for the event of interest is accomplished through analytical, numerical or Monte Carlo techniques.

在许多问题中，信息可以从许多不同的来源获得。汇总这些不同的信息对于良好的决策非常重要，但在开发描述信息特征的模型时也会导致特殊的统计问题。这方面的先前研究主要依赖于使用历史数据作为信息源建模的基础。提出的研究发展了一个贝叶斯框架，决策者可以使用它来编码他对信息源的主观知识，以便汇总他们的信息。该框架具有高度灵活的环境，用于对信息源的概率性质和相互关系进行建模，并要求使用经过验证的决策分析评估技术进行直接和直观的主观判断。通过分析、数值或蒙特卡罗技术对评估结果进行分析，以产生感兴趣事件的后验分布。