AVIATION ACCIDENT AND INCIDENT FORECASTING COMBINING OCCURRENCE INVESTIGATION AND METEOROLOGICAL DATA USING MACHINE LEARNING

Studies on safety in aviation are necessary for the development of new technologies to forecast and prevent aeronautical accidents incidents. When predicting these occurrences, literature frequently considers internal characteristics operations, such as aircraft telemetry flight procedures, or external characteristics, meteorological conditions, with only few relationships being identified between two. In this study, data from 6,188 occurrences involving accidents, incidents, serious Brazil January 2010 October 2021, well two automatic weather stations, totaling more than 2.8 million observations, were investigated using machine learning tools. For analysis, decision tree, extra trees, Gaussian naive Bayes, gradient boosting, k-nearest neighbor classifiers a high identification accuracy 96.20% used. Consequently, developed algorithm can predict functions operational patterns. Variables maximum take-off weight, registration model, wind direction among main forecasters This study provides insight into measures occurrences.