Modeling the Uncertainty Due to Data/Visual Transformations Using Sensitivity Analysis

In this project, the investigators study the fundamental aspects ofincorporating uncertainty with sensitivity analysis in the visualanalytics process. They also aim to develop novel and scalablevisual representations of sensitivity, from the visualization of theraw sensitivity coefficients to visual summaries of multivariatederivatives obtained from the analysis. Uncertainty-aware visualanalytics helps enhance analysts' confidence levels on the insightgained from the analysis. Furthermore, it gives toolmakers amethodology for measuring and comparing the robustness ofdata and visual transformations. Sensitivity coefficients of data andvisual transformations are useful for discovering the factors that mostlycontribute to output variability, identifying stability regions of thedifferent transformations within the original data space, andtelling the analyst what the interaction is between variables,outputs and transformations.

Uncertainty is introduced throughout the process of data generation,transformation, and analysis in most real-world applications. The ability toincorporate uncertainty into visual analysis is therefore critically importantfor insightful reasoning and key decision making. This project will have awide-reaching impact on those areas relying on the ability to reasonabout large amounts of data. On one hand, it suggests to provide a variationalview of the visual analytics process, which opens up new directions andparadigms for visual data analysis and mining. On the other hand,the improved understanding of the visual analytics process will helpestablish the field as a scientific discipline.