抄録
One of the precious applications of Genetic Algorithms (Gas) is its application to multi-objective optimization (MO). In MO, we need a number of scalar optimization to obtain whole sets of Pareto solution for adequate information for decision-making. Thus, it becomes very time consuming and difficult to precede MO. In the proposed method, we use generalized data envelopment analysis (GDEA) for estimation of Pareto optimality. In GDEA, we can obtain Pareto sets even if it has some non-convex feature. Important key to obtain whole Pareto sets, we would like to concentrate searching range between interval of Pareto sets. In this paper, we will propose a new type of adaptive range genetic algorithms for that purpose. We will investigate the efficiency of the method, through numerical examples.
