Collaborative Research on Dependence Assessment for Decision and Risk Analysis

在与杜克大学的Robert Winkler的合作下，这个PI提议开发使用联结法的程序来评估依赖性。这种方法将被应用的情况是专家知识，其中决策者必须考虑一组相互关联的随机变量。这项研究将包括对连词和依赖度量的数学发展和分析、评估程序的设计、测试专家理解这些程序的能力的实验室研究、计算程序的开发、实际知识建模应用的实地研究，以及依赖性评估的协议和一般规定性建议的构建。

In collaboration with Robert Winkler of Duke University, this PI proposes to develop procedures using the copula approach in order to assess dependence. The situations in which this approach will be applied are expert knowledge ones in which the decision maker must consider a set of interrelated random variables. The research will involve the mathematical development and analysis of families of copulas and dependence measures, the design of assessment procedures, laboratory studies to test the ability of experts to understand the procedures, development of computational procedures, field studies with real-world knowledge-modeling applications, and the construction of protocols and general prescriptive recommendations for dependence assessment.