Visual Analytics based Search-Analyze-Forecast Framework for Epidemiological Time-series Data

tuna gonen, Yiwen Xing, Cagatay Turkay, Alfie Abdul-Rahman, Radu Jianu, Hui Fang, Euan Freeman, Franck P Vidal, Min Chen

Room: 104

2023-10-21T22:00:00ZGMT-0600Change your timezone on the schedule page
Exemplar figure, but none was provided by the authors

The COVID-19 pandemic has been a period where time-series of disease statistics, such as the number of cases or vaccinations, have been intensively used by public health professionals to estimate how their region compares to others and estimate what future could look like at home. Conventional visualizations are often limited in terms of advanced comparative features and in supporting forecasting systematically. This paper presents a visual analytics approach to support data-driven prediction based on a search-analyze-predict process comprising a multi-metric, multi-criteria time-series search method and a data-driven prediction technique. These are supported by a visualization framework for the comprehensive comparison of multiple time-series. We inform the design of our approach by getting iterative feedback from public health experts globally, and evaluate it both quantitatively and qualitatively.