Aircraft Damage Classification by using Machine Learning Methods

Abstract

In the general concept of aviation safety comes first, so country authorities and airlines have focused on increasing aviation safety. The introduction of cockpit voice recorder and flight data recorder, crew resource management (following the Tenerife disaster), the airborne collision avoidance system (ACAS), and ground proximity warning systems (GPWS) are the examples of specific improvements for aircraft’s configuration because of various accidents (ICAO, 2013c). Due to the increasing trend of air transportation, the rate had risen to approximately 3 accidents in 1 million departures to approximately 50 accidents in 1 million departures in 1960 and fell approximately 5 accidents in 1 million departures in the 1980s (Boeing, 2016). The International Air Transport Association (IATA) stated commercial aircraft carriers, which already provide the transportation process of more than 3 billion people across the world each year in 2019 (IATA, 2019; Zhang & Mahadevan, 2019). To prevent aviation accidents, it is crucial to understand how minor mishaps turn into major catastrophes by specifying the processes about abnormal occurrences spread throughout the system. Therefore, the safety of air transportation is impacted by several issues such as passenger number, temperature, dew point, wind speed, wind direction, latitude, longitude, and time-periods as time, month, and year (Alle et al., 2009)