Description and First Evaluation of an Approach for a Pilot Decision Support System Based on Multi-attribute Decision Making

Multi-attribute decision making methods often rely on complicated mechanisms to represent the preferences of the decision makers. The need to employ these mechanisms coupled with the effort to come up with suitable preferences means that these methods are used sparingly and only in important decisions. However, often times, decision makers need to make choices under time pressures and without the possibility to consider their scenario in exhausting detail. In this paper a new multi-attribute decision-making approach is presented. The main idea is to have a single solution recommended to the decision maker from the Pareto front of decisions. The recommendation mechanism is based on using a weighted sum approach on the attributes of the decisions. The idea is to weigh the attributes so that averaged over multiple scenarios, where a similar decision needs to be made, the values for the objectives would be as low as possible. This paper serves as a showcase of this approach, framing it within the context of a possible centerpiece of a system for dynamic alternate airport selection. The approach was examined on a custom data set and employed the NSGA-II algorithm to search for the weights. The approach shows promising results that could be improved by using a more sophisticated selection method.