A hybrid multiobjective immune algorithm with region preference for decision makers

Licheng Jiao,Wei Zhang,Ruochen Liu,Fang Liu

DOI: https://doi.org/10.1109/CEC.2010.5586345

2010-01-01

Abstract:Recently, one of the main tools of decision maker (DM) preference incorporation in the multiobjective optimization (MOO) has been using reference points and achievement scalarizing functions (ASF). The core idea of these methods is converting the original multiobjective problem (MOP) into single objective problem by using ASF to find a single preferred point. However, many DMs not only interest in a single point but also a set of efficient points in their preferred region. In this paper, we introduce a hybrid multiobjective immune algorithm (HMIA) for DM. It combines the immune inspired algorithm and region preference based on a novel dominance concept called region-dominance without ASF. The new algorithm can let DMs flexibly decide the number of reference points and accurately determine the preferred region with its simple and effective interactive methods. To exemplify its advantages, simulated results of HMIA are shown with some well-known problems.