Uncertainty Visualization

Abstract

We proposed a novel design for uncertainty visualization, which helps users interpret uncertainty generated by predictive models and make informed decisions. We presented three candidate visualizations and evaluated them in a crowd-sourced study via Amazon Merchanical Turk (AMT). The results show that participants using our design feel more confident and make more accurate estimations. Furthermore, we discuss the benefits of our design in visualization for uncertainty visualizations.

Key Words: Interactive uncertainty visualization, uncertainty interpretation, descriptive plot, descriptive statistics, probability distribution

Paper is under review. More information about this project will be disclosed soon!