VISPLORE

Namrata Khemka, Christian J. Jacob · 2009

VISPLORE is an interactive toolkit to visualize data generated from population-based optimization algorithms. In particular, we demonstrate VISPLORE's capabilities by exploring solutions from particle swarm optimization on different levels - from individual solutions, to populations (as sets of solutions), to experiments (as sets of populations), and to collections of experiments. Users can control aspects of the various visual representations to view multi-dimensional data produced over time. Furthermore, our application includes a large range of automatic skimming tools, controlled by manual and automated sliders, and supports interactive manipulations. By using dynamic visualization techniques, we provide instant visualizations customizable by the user for data exploration tasks.

Read the paper · More papers on PaperTik