Cross validation was used for model development, while nested cross validation was used to validate the models. The model found to have the best performance in predicting air traffic controller acceptance or rejection of a route change, using the available data from Fort Worth Air Traffic Control Center and its adjacent Centers, was the random forest, with an F-score of 0.77. This result indicates that the operational acceptance of reroute requests does indeed have some level of predictability, and that, with suitable data, models can be trained to predict the operational acceptability of reroute requests. The predictor was developed using data mining techniques applied to flight plan amendment data and data from a trial of the NASA developed DWR tool at American Airlines in 2014. Routes are classed as operationally unacceptable to ATC if they are either modified by the air traffic controller before being implemented in the form of a Center route amendment, or do not result in any Center route amendment being implemented at all. Features identified as relevant include (1) the historical usage of the route change, (2) the proximity of the reroute start point to the boundaries of the airspace sector containing the reroute start, (3) the length of the reroute, and (4) the congestion in the reroute start sector.
The application of data mining to the DWR trial data presented in this paper provides significant results indicating that the operational acceptance of reroute requests is indeed predictable, and that, with suitable data, models can be trained to predict the operational acceptability of reroute requests.
Collected and summarized from the source below by Minh Pham https://db.vista.gov.vn:2095/science/article/pii/S0968090X18313901