Doctoral Dissertation Research in DRMS: Cross-Disciplinary Analyses Using Multi-Attribute Utility Theory

这项工作的目标是研究三种特定类型的决策，涉及一个以上的目标。 第一组决定包括那些与健康结果有关的决定。 这些可以是医疗决定，也可以是与生活方式选择有关的决定。 这些决定往往涉及到寿命长短和整体福祉之间的权衡。 第二组决定包括那些在地理上界定结果的决定。 这是环境决策的一个典型特征。 这些决策涉及对许多不同地区的一个或多个因素进行评估，并且通常过于复杂，如果没有决策分析工具的帮助，实际上无法执行。 第三组决策包括在竞争博弈论的情况下做出的决策。 传统的博弈论通常只考虑这些情况下的物理或货币结果，但决策分析可以很好地研究其他因素对参与者的影响，这三种决策类型的共同特征是，每一种都涉及试图同时最大化或最小化多个目标。 要对这些决策进行可行的分析，必须对决策者对可能结果的偏好结构做出某些假设。 我们的目标是开发足够简单的模型，使决策易于处理，但足够详细，以准确地反映偏好。 在所研究的所有三种情况下，更好地理解如何做出这些决定将是非常有益的。

The goal of this work is to study three particular types of decisions that involve more than one objective. The first set of decisions includes those with health outcomes. These can be either medical decisions, or decisions relating to lifestyle choices. These decisions often involve tradeoffs between length of life and overall well-being. The second set of decisions includes those with outcomes that are defined geographically. This is a typical feature of environmental policy decisions. These decisions involve evaluating one or more factors over many different regions, and are often too complex to be carried out practically without the help of decision analysis tools. The third set of decisions include those made in competitive game theory situations. Traditional game theory often considers only the physical or monetary outcome of these situations, but decision analysis is well-equipped to study the effects of other factors which may be important to the players.The common feature shared by these three decision types is that each one involves trying to maximize or minimize multiple objectives simultaneously. For these decisions to be feasibly analyzed, certain assumptions must be made about the structure of decision makers' preferences over the possible outcomes. The goal is to develop models simple enough to make the decisions tractable, but detailed enough to accurately reflect the preferences. In all three situations being studied, a better understanding of how these decisions should be made would be extremely beneficial.