Epstein Department of Industrial and Systems Engineering, Price School of Public Policy, University of Southern California, Los Angeles, California 90089, USA
Dr. Ali Abbas
Epstein Department of Industrial and Systems Engineering, Price School of Public Policy, University of Southern California, Los Angeles, California 90089, USA
Multiattribute utility theory has a long history of application in engineering and systems design. These applications rely almost exclusively on two types of utility functions: the additive and multiplicative forms. The foundations of utility theory, however, do not place restrictions on the functional form of the utility function, meaning that many methods of representing preferences remain unexplored in systems design. This paper reviews the literature on applications of utility theory in design to bring light to potential new directions for research and to clarify a few subtle misapplications of utility theory. These misuses include the failure to distinguish between direct and indirect value attributes, the use of probability as an attribute, treating costs from different sources differently, and restricting the functional form of utility. This paper introduces a value-based approach that is based on the creation of a deterministic value function for design and the assignment of a 1-D utility function over the value measure. We use a conceptual and numeric example to illustrate the greater flexibility of this approach. We also present the concept of utility transversality in engineering design. We show that several criticisms of utility theory that have appeared in the engineering design literature are actually criticisms of these artificial limitations and that these limitations are overcome by the value-based approach.
Entropy methods enable a convenient general approach to providing a probability distribution with partial information. The minimum cross-entropy principle selects the distribution that minimizes the Kullback–Leibler divergence subject to the given constraints. This general principle encompasses a wide variety of distributions, and generalizes other methods that have been proposed independently. There remains, however, some confusion about the breadth of entropy methods in the literature. In particular, the asymmetry of the Kullback–Leibler divergence provides two important special cases when the target distribution is uniform: the maximum entropy method and the maximum log-probability method. This paper compares the performance of both methods under a variety of conditions. We also examine a generalized maximum log-probability method as a further demonstration of the generality of the entropy approach.
Archimedean utility copulas comprise the general class of multiattribute utility functions that have additive ordinal preferences and are strictly increasing with each argument for at least one reference value of the complementary attributes. The construction of an Archimedean utility copula requires an assessment of an individual utility function for each attribute as well as a single generating function. The assessment of individual utility functions for the attributes of a decision has had a large share of literature coverage, but there has been much less literature on the construction of the generating function for the Archimedean functional form. This paper focuses on the assessment of Archimedean utility copulas with polynomial generating functions. We provide methods to assess these generating functions and derive bounds on the types of utility surfaces that they provide. We demonstrate that linear generating functions correspond to the multiplicative form of mutual utility independence, and then we show how higher-order polynomial generating functions allow more flexibility in the types of multiattribute utility functions and corner values that can be modeled. The results of this paper provide a new method to help the analyst construct multiattribute utility functions in a simple way when utility independence conditions do not hold.
Winner of the 2025 Clemen–Kleinmuntz Decision Analysis Best Paper Award The value of information is an important concept in decision analysis that has been quantified as the buying price (BPI) for the information and the expected utility increase (EUI) obtainable by using the information. These two measures rank information sources identically in a scalar-valued decision problem only when the utility function is linear or exponential. In contrast, this paper focuses on the value of information across scalar-valued decision problems sharing the same utility function such as different divisions within an organization exploring various information sources for their decisions using the same organizational utility function. In this context, it still makes sense to ask which sources are more informative. We show that BPI and EUI rank information sources identically in this context only when the utility function is linear. However, if the certainty equivalent increase is used instead of EUI, then identical ranking with BPI across problems is maintained for the broader class of linear or exponential utility functions. We discuss the importance of these results for distributed decision-making settings, where different departments within an organization may calculate the value of information separately. Our results advise against using EUI to measure information value in this context when risk attitude is important.