Abstract (eng)
In disaster management, decisions have to be fair in order to be accepted by the majority of the population. In this work, shelters are built before and after a disaster occurred and the locations of these shelters are chosen so that fairness is maximized. The fairness of a possible solution is evaluated by two different approaches, ex-ante and ex-post, that are visualized using a two-stage decision tree. Small problem instances are generated, first solved by complete enumeration, and then compared with the results of a genetic algorithm, while large problem instances are only solved by the genetic algorithm due to otherwise excessive computation times. In addition, the solutions of the ex-ante/ex-post objective function are plugged into the ex-post/ex-ante objective function and the percentage differences are compared. Afterwards, the ex-ante and ex-post solutions are plugged into the utilitarian objective function and compared with the utilitarian solution.