Development of Rational Decision Making Procedure Based on Additive Multi-Attribute Utility Function Using Strict Preference Information

摘要

A utility function consists of multiple attributes where the scaling constants corresponding to the attributes describe a decision making problems to select the most preferred alternative for decision maker. In the conventional method such as Keeney and Raiffa(1976), contradiction is sometime included if the burden by the question to the decision maker is large. This study proposes a new method to estimate scaling constants based on strict preference information which is obtained on the answer of the questions that are relatively less burdensome than existing procedures of multi attribute utility analysis. Numerical experiments are conducted on the decision making problem about agricultural policy and it indicates usefulness of the proposed procedure.