Pareto-RadVis: A Novel Visualization Scheme for Many-objective Optimization
Mahda Nasrolahzadeh, Amin Ibrahim, Shahryar Rahnamayan, Javad Haddadnia · 2020
Interest in visual data analytics related to many-objective optimization has recently risen. This paper introduces a novel visualization scheme based on the Radial Coordinate Visualization (RadVis) for analysis of Pareto fronts during the optimization process. This method illustrates the ranks of the Pareto front, the relative location, and the distribution of candidate solutions. The results show that the proposed method is capable of showing different ranks of Pareto fronts simultaneously. The simplicity of the P-RadVis visualization and its compatibility to work with many-objective algorithms could be beneficial in terms of visual analytics for real-time monitoring of optimization process.