Robust Assessment of Preference Functions
A framework for assessing a single decision maker’s preference function of several variables is sketched. The preference function is assumed to be additively decomposable into onedimensional preference functions. All attributes are known prior to the given analysis. The case of probability distributions can basically be dealt with in the same way as the case of certainty. However, in the first case we explore an instability phenomenon which does not exist for sure alternatives. The approach is applied to a real world problem in environmental decision making. Preferences serve as a proxy measure for unobtainable statistical data on damage cost and frequency. We describe this application along an outline of a software system developed to cope with that problem.
KeywordsUtility Function Preference Function Strict Preference Preference Elicitation Hazard Class
Unable to display preview. Download preview PDF.
- CPLEX Optimization Inc. (1989), ”Using the CPLEX™ linear optimizer”.Google Scholar
- Despotis, D.K., Yannacopoulos, D., and Zopounidis, C. (1990), ”A review of the UTA multicriteria method and some improvements”, Foundations of Computing and Decision Science 15, 63–76.Google Scholar
- Kämpke, T. (1992), Bestimmung von Präferenzfunktionen mittels linearer und nicht- linearer Optimierungsverfahren, in progress.Google Scholar
- Kämpke, T., Radermacher, F.J., and Wolf, P. (1993), ”Supporting preference elicitation”, Decision Support Systems, to appear.Google Scholar
- Keeney, R.L. and Raiffa, H. (1976), Decisions with multiple objectives, Wiley, New York.Google Scholar
- Wolf, P. (1992), Rechnerunterstützte Elizitierung mehrattributiver Präferenzstrukturen, Dissertation, University of Ulm.Google Scholar