Visual Analytics of Multidimensional Dynamic Data with a Financial Case Study

This work deals with a problem of analysis of time variant objects. Each object is characterized by a set of numerical parameters. The visualization method is used to conduct the analysis. Insights of interest for the analyst about the considered objects are obtained in several steps. At the first step, a geometric interpretation of the initial data is introduced. Then, the introduced geometrical model undergoes several transformations. These transformations correspond to solving the first problem of the visualization method, in particular, obtaining visual representations of data. The next step for the analyst is to analyze the generated visual images and to interpret the results in terms of the considered objects. We propose an algorithm for the problem solution. Developed interactive visualization software is described, which implements the proposed algorithm. We demonstrate how with this software the user can obtain insights regarding the creation and disappearance of object clusters and bunches and find invariants in the initial data changes.