An Entropy Approach for Utility Assignment in Decision Analysis

BAYESIAN INFERENCE AND MAXIMUM ENTROPY METHODS IN SCIENCE AND ENGINEERING: 22nd International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering

The author: Ali E. Abbas
Department of Management Science and Engineering, Stanford University, Stanford, Ca, 94305

Abstract

A fundamental step in decision analysis is the elicitation of the decision-maker's preferences about the prospects of a decision situation in the form of utility values. However, this can be a difficult task to perform in practice as the number of prospects may be large, and eliciting a utility value for each prospect may be a time consuming and stressful task for the decision maker. To relieve some of the burden of this task, this paper presents a normative method to assign unbiased utility values when only incomplete preference information is available about the decision maker. We introduce the notion of a utility density function and propose a maximum entropy utility principle for utility assignment. ©2003 American Institute of Physics