Bayesian interactive decision support for multi-attribute problems with even swaps
Abstract
Even swaps (ES) is a multi-criteria decision-making method introduced by Hammond et al.(1998) that makes it easier for decision makers (DMs) to make trade-offs between the decisioncriteria. The ES method can be further guided using decision support systems (DSSs) such that itbecomes even easier to use the method.This thesis intends to make a DSS to guide a DM through the ES method and assess how thepreferences of the DM can be captured and updated using probabilistic dominance based onBayesian updating.Results show that the DSS implemented in this thesis can remove dominated alternatives throughabsolute dominance and practical dominance. Furthermore, the DM can make ES through theDSS. In addition, the DSS can suggest alternatives that are likely to be dominated, while alsosuggesting objectives that are close to having equal ranks such that these alternatives andobjectives can be used for ES. Finally, changing the coordinate pair and lower and upper limitsof the uniform distribution produces different results for the probable dominance such that itbecomes easy to see which coordinate pair and limits for the uniform distribution produces thebest results for dealing with the preferences of the DM.