Explainable Boosting Machine for Predicting Wind Shear-Induced Aircraft Go-around based on Pilot Reports

Afaq Khattak,Pak-wai Chan,Feng Chen,Haorong Peng
DOI: https://doi.org/10.1007/s12205-023-0410-8
IF: 2.115
2023-09-20
KSCE Journal of Civil Engineering
Abstract:The go-around is a safety-critical procedure in civil aviation that is rarely executed but is essential to avoid risky landings. Analyzing the factors that trigger go-around events can aid in identifying measures that could lower their frequency. This involves circumstances that could be deemed abnormal and intrinsically harmful. The study employed the Explainable Boosting Machine (EBM), a contemporary transparent model, to predict aircraft go-arounds and interpret different influential factors. The model proposed exhibits comparable accuracy to black-box models. The study utilized pilot reports and applied SMOTE-ENN to address the imbalance problem. The EBM model was trained with treated data in conjunction with Bayesian optimization. The EBM model's performance was evaluated using holdout data and compared to binary logistic regression and decision tree models, as well as black-box models such as adaptive boosting, random forest, and extreme gradient boosting. The EBM model exhibited superior performance compared to other models in terms of precision (83.15%), recall (79.77%), geometric mean (77.29%), and Matthews's correlation coefficient (0.453). The EBM algorithm enables the comprehensive interpretation of individual and pairwise factor interactions in predicting aircraft go-around outcomes from both global and local perspectives. This facilitates the assessment of the impact of different factors on go-around outcomes.
engineering, civil
What problem does this paper attempt to address?