| Abstract | This paper presents an approach for constructingvisual representations of high dimensional objective spacesusing virtual reality. These spaces arise from the solutionof multi-objective optimization problems with more than 3objective functions which lead to high dimensional Pareto frontswhich are difficult to use. This approach is preliminarily investigatedusing both theoretically derived high dimensional Paretofronts for a test problem (DTLZ2) and practically obtainedobjective spaces for the 4 dimensional knapsack problem viamulti-objective evolutionary algorithms like HLGA, NSGA, andVEGA. The expected characteristics of the high dimensionalfronts in terms of relative sizes, sequencing, embedding andasymmetry were systematically observed in the constructedvirtual reality spaces. |
|---|
