Keywords: data visualization, interactive, dashboard, statistical modeling, model diagnostics
IHME produces data visualizations for an external global health audience that are populated by the results of complex data processing and statistical modeling pipelines. Researchers and statistical modelers need to have access to intermediate results and diagnostics throughout these pipelines to ensure their accuracy. This presentation focuses on the case study of building an interactive visualization to view intermediate results and diagnostics for the cause of death prediction pipeline at IHME.