The capabilities of the envisioned performance management services of the NextGen may offer means for testing of measures of human performance derived from operational data and allow for longitudinal studies on the effects of operational NextGen on human operators. Time-based metrics offer an attractive solution to measurement challenges of the NextGen, given the dynamic nature of the system, its dependence on severe time constraints, and the role of time in critical aspects of human performance, mainly workload and situation awareness. Due to potentially very high number of variables and complexity of the underlying data structure that render standard statistical techniques inadequate, novel techniques that perform nonlinear regression and pattern recognition along with feature selection and variable selection are potential candidates for such analyses. Promising statistical techniques to handle the many problems associated with these kinds of data include Generalized Linear and Generalized Linear Mixed Models, and Support Vector Machines.
& Fokoue, E.
(2011). Data Mining Techniques to Derive Human and System Performance Measures of Air Traffic Control from Operational Data. 16th International Symposium on Aviation Psychology, 529-534.