Visualization-based Multi-Criterion Decision Making with NIMBUS Method Using PaletteViz

After many efficient evolutionary many-objective optimization (EMaO) algorithms have been proposed and demonstrated to find multiple well-distributed near Pareto-optimal solutions, one main emphasis now is to combine them with suitable multi-criterion decision-making (MCDM) approaches to choose a single preferred solution. For integrating MCDM approaches with EMaO algorithms, there have been growing interests in implementing MCDM concepts algorithmically within EMaOs, but in this paper, we propose a visualization-based MCDM-EMaO integration implementing the well-known NIMBUS method using a recently proposed PaletteViz visualization technique and demonstrate its working by applying it to two test problems and an engineering design problem. The detailed results show the usefulness of the PaletteViz procedure in assisting decision-makers to implement MCDM approaches with a better understanding of trade-off solutions.