US-United Kingdom Cooperative Science: Decision and Risk Analysis

拟议的研究项目是最初合作努力的结果，旨在审查和评估决策和风险分析领域的知识状况，目的是为这些学科的未来研究方向提供一个视角。最近对这些领域的发展和状况的审查表明了四个广泛而重要的趋势：(1)增加对决策和风险分析应用的关注；(2)规范性和描述性研究人员在直接和相关学科之间进行更多的跨学科研究；(3)不断努力重新审视现有理论；(4)用于决策支持和培训的交互式智能软件加速发展。与这些广泛的研究趋势相一致，本建议的研究项目分为以下两类：(1)理论和方法的发展；(2)风险分析和管理。要解决的理论和方法问题包括模型制定、建模不确定性、建模偏好和优化。所提出的工作的一个基本前提是，通常获得的信息比达到最佳选择所需的信息要多。此外，由于来自个人的精确输入通常是不可靠的，不精确的信息可能更相关、更合适，而且是做出理性决策所必需的一切。在此前提下，研究者需要重新审视现有的理论，构建和测试（从计算和行为两方面）交互、集成、智能的决策支持模型，以促进符合此前提的模型构建、不确定性和偏好评估以及优化。研究结果应该提供关于人类推理和决策能力和行为的新的规范和描述性知识，新的方法，以及支持决策培训和实践的智能交互式软件。这项研究将与伦敦商学院的Derek W. Bunn教授合作进行。美英合作科学计划的目标是促进互利的研究合作，使美国科学家能够利用特殊的研究机会。该奖项实现了这一目标，使赫伯特·莫斯科维茨博士能够与具有互补专业知识的英国科学家合作三年。

The proposed research project is an outgrowth of an initial collaborative effort to review and assess the state of knowledge in the fields of decision and risk analysis, with the aim of providing a perspective for future research directions in these disciplines. A recent review of the evolution and state of the fields suggests four broad and significant trends: (1) increased focus on decision and risk analysis applications, (2) more interdisciplinary research among normative and descriptive researchers in direct and allied disciplines, (3) growing efforts to reexamine existing theory, and (4) accelerated growth of interactive, intelligent software for decision support and training. Consistent with these broad research trends, this proposed research project is divided into the following two categories: (1) Theory and methodological development and (2) risk analysis and management. Theory and methodological issues to be addressed include model formulation, modeling uncertainty, modeling preferences, and optimization. A fundamental premise in the proposed work is that usually, more information is acquired than is necessary to reach an optimal choice. Moreover, as precise inputs from individuals are generally unreliable, imprecise information may be more relevant, appropriate, and all that is required to arrive at a rational decision. Given this premise, the investigators shall reexamine the existing theory and build and test (both computationally and behaviorally) interactive, integrated, intelligent decision support models to facilitate model building, uncertainty and preference assessment, and optimization consistent with this premise. The results of the research should provide new normative and descriptive knowledge regarding human inference and decision making capabilities and behavior, new methodologies, and intelligent, interactive software to support decision making training and practice. This research will be carried out in collaboration with Professor Derek W. Bunn of the London Business School. It is the objective of the U.S.-United Kingdom Cooperative Science Program to promote mutually beneficial research cooperation and to enable U.S. scientists to take advantage of special research opportunities. This award fulfills that objective by enabling Dr. Herbert Moskowitz to spend three years in collaboration with British scientists with complementary expertise.