Task Structure of evaluation analysis of aberrancy detection


Model of aberrancy detection as part of evaluation analysis
Using task analytic approach, we have identified subtasks of evaluating aberrancy detection algorithms: Obtain Surveillance Data, Obtain True Outbreak Data, Monitor Data for Aberrancies (by applying an Aberrancy Detection Method being evaluated), Stratify Monitoring Results, Measure Performance (by comparing the results of aberrancy detection to a gold standard using a performance metric of interest), and Measure Aggregate Performance (allowing for the analysis of trade-offs among different performance metrics under different settings).

← Back to Models