33th Congress of the International Council of the Aeronautical Sciences

09 - Air Transport System Efficiency

MACHINE LEARNING METHODS ENSURING BOTH PERFORMANCE AND INTERPRETABILITY OF ESTIMATING AIRCRAFT ARRIVAL TIMES

N. Morikawa, Faculty of Aerospace Engineering, The University of Tokyo, Japan; E. Itoh, Air Traffic Management Department, Electronic Navigation Researc, Japan

To further improve efficiency of runway management, we applied machine learning models to estimate aircraft arrival times using real data in Japan. Various indicators such as SHAP values were introduced to ensure not only the modelsí performance, but also interpretability for stakeholders involved in the operation.


View Paper