Analysis of aircraft accidents and incidents using artificial intelligence

This thesis examines how artificial intelligence methods can be applied to the analysis of aircraft accidents and incidents. The study first addresses the fundamental causes of aviation accidents in detail, including human factors, technical failures, weather conditions, organizational factors, and external influences. Subsequently, accident report data were utilized to conduct a word frequency analysis, identifying frequently recurring concepts and examining the relationships between these concepts. In the methodology section, processes such as pre-processing, n-gram analysis, frequency calculation methods, and the contribution of artificial intelligence are explained in detail. Furthermore, a logistic regression-based artificial intelligence model was developed to estimate the likelihood of accident occurrence. The outputs of the model revealed which factors were more strongly associated with accidents. The findings contribute to identifying preventive measures for reducing aircraft accidents. In particular, recommendations were made regarding the improvement of maintenance processes, training programs addressing human factors, and operational decisions related to weather conditions. In conclusion, this thesis demonstrates the applicability of artificial intelligence methods in the field of aviation safety and provides significant findings that may guide future research.