Bayesian Estimation of the Weights of Multi-Criteria Decision Alternatives Using Monte Carlo Integrations
Nov 30 1989
A Bayesian procedure is proposed for the estimation of the weights in a multi-criteria decision model with data that stem from pairwise comparison of preferences. The prior information implies that the weights are restricted by means of inequality restrictions to the unit simplex. The posterior results are computed by Monte Carlo integration procedures. The Bayesian procedure is applied to a case study, described by Lootsma (1980), concerning the choice of a professor in Operations Research (OR). Results of our case study are: (1) according to the Bayesian procedure a different candidate would be chosen as professor of OR than according to the non-Bayesian procedure; (2) given the available prior and data information, there exists a substantial probability of taking the wrong decision.