Explainable prediction of Qcodes for NOTAMs using column generation

A NOtice To AirMen (NOTAM) contains important flight route related information. search and filter them, NOTAMs are grouped into categories called QCodes. In this paper, we develop a tool to predict, with some explanations, Qcode for NOTAM. We present way extend the interpretable binary classification using column generation proposed in Dash, Gunluk, Wei (2018 S., O., & Wei, D. (2018). Boolean decision rules via generation. S. Bengio, H. Wallach, Larochelle, K. Grauman, N. Cesa-Bianchi, R. Garnett (Eds.), Advances neural information processing systems. (Vol. 31) Curran Associates, Inc. Retrieved from https://proceedings.neurips.cc/paper/2018/file/743 394beff4b1282ba735e5e3723ed74-Paper.pdf [Google Scholar]) multiclass text method. describe techniques used tackle issues one-vs-rest classification, such as multiple outputs class imbalances. Furthermore, introduce heuristics, including use of CP-SAT solver subproblems, reduce training time. Finally, show that our approach compares favorably state-of-the-art machine learning algorithms like Linear SVM small networks while adding needed interpretability component.