Optimization of utility and value functions

For more than a decade, multiattribute utility/value theory and multiobjective mathematical programming have offered different approaches to similar problems. Unfortunately, the two areas have developed with little interaction in spite of their common aims. This paper considers the use of utility/value functions in a mathematical programming framework, and demonstrates that these functions often possess desirable properties from an optimization point of view. It concludes that a hybridization of approaches is more viable than is perhaps commonly assumed.