Visualizing swarm behavior with a particle density map
Hyeon-Chang Lee, Yong-Hyuk Kim · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2019
In general, when a problem of high dimension is solved with particle swarm optimization (PSO), a large number of particles is used. However, it is quite difficult to understand this method's search process. To address this, the best known solutions of PSO were fixed as the center point, after which the solutions were reduced to 2 dimensions using Sammon mapping and the search process was visualized using a heatmap. As a result, the PSO process of searching for various optimization functions was able to be understood intuitively, and PSOs possessing different parameters could be compared