Depth-Based Visualizations for Ensemble Data and Graphs
Abstract: Ensemble data sets appear in many domains. Ensemble visualization is a way to study the underlying generating model of data by analyzing ensembles of solutions or measurements. The objective of ensemble visualization is often to convey characteristics of typical/central members, outliers, and variability in the ensemble. In the absence of any information about the generative model, a family of nonparametric methods, known as data depth, provides a quantitative notion of centrality among ensemble members. Data depth methods also form the basis of several ensemble visualization techniques, including the popular Tukey boxplot.
In this talk I will describe my work in developing novel data depth-based visualizations and their advantages over existing visualization techniques, for various types of data, namely, ensembles of three-dimensional isocontours, ensembles of paths on a graph, ensemble data in high-dimensional spaces, and graphs.