{"title":"基于集成机器学习多类分类模型和马尔可夫链的座舱乘员安全性能预测","authors":"Naimeh Borjalilu, Fariborz Jolai, Mahdieh Tavakoli","doi":"10.3846/aviation.2023.19739","DOIUrl":null,"url":null,"abstract":"The main tool of cockpit crew performance evaluation is the recorded flight data used for flight operations safety improvement since all certified airlines require implementation of a safety and quality management system. The safety performance of a flight has been a challenging issue in the aviation industry and plays an important role to acquire competitive benefits. In this study, an integrated multi-class classification machine learning models and Markov chain were developed for cockpit crew performance evaluation during their flights. At the outset, the main features related to a flight are identified based on the literature review, flight operations expert’s statements, and the case study dataset (as numerical example). Afterwards, the flights’ performance is evaluated as a target column based on four multi-class classification models (Decision Tree, Support Vector Machine, Neural Network, and Random Forest). The results showed that the random forest classifier has the greatest value in all evaluation metrics (i.e., accuracy = 0.90, precision = 0.91, recall = 0.97, and F1-score = 0.93). Therefore, this model can be used by the airline companies to predict flight crew performance before the flight in order to prevent or decrease flight safety risks.","PeriodicalId":51910,"journal":{"name":"Aviation","volume":"125 1","pages":"0"},"PeriodicalIF":0.8000,"publicationDate":"2023-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"COCKPIT CREW SAFETY PERFORMANCE PREDICTION BASED ON THE INTEGRATED MACHINE LEARNING MULTI-CLASS CLASSIFICATION MODELS AND MARKOV CHAIN\",\"authors\":\"Naimeh Borjalilu, Fariborz Jolai, Mahdieh Tavakoli\",\"doi\":\"10.3846/aviation.2023.19739\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The main tool of cockpit crew performance evaluation is the recorded flight data used for flight operations safety improvement since all certified airlines require implementation of a safety and quality management system. The safety performance of a flight has been a challenging issue in the aviation industry and plays an important role to acquire competitive benefits. In this study, an integrated multi-class classification machine learning models and Markov chain were developed for cockpit crew performance evaluation during their flights. At the outset, the main features related to a flight are identified based on the literature review, flight operations expert’s statements, and the case study dataset (as numerical example). Afterwards, the flights’ performance is evaluated as a target column based on four multi-class classification models (Decision Tree, Support Vector Machine, Neural Network, and Random Forest). The results showed that the random forest classifier has the greatest value in all evaluation metrics (i.e., accuracy = 0.90, precision = 0.91, recall = 0.97, and F1-score = 0.93). Therefore, this model can be used by the airline companies to predict flight crew performance before the flight in order to prevent or decrease flight safety risks.\",\"PeriodicalId\":51910,\"journal\":{\"name\":\"Aviation\",\"volume\":\"125 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.8000,\"publicationDate\":\"2023-10-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Aviation\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.3846/aviation.2023.19739\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"ENGINEERING, AEROSPACE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Aviation","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3846/aviation.2023.19739","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, AEROSPACE","Score":null,"Total":0}
COCKPIT CREW SAFETY PERFORMANCE PREDICTION BASED ON THE INTEGRATED MACHINE LEARNING MULTI-CLASS CLASSIFICATION MODELS AND MARKOV CHAIN
The main tool of cockpit crew performance evaluation is the recorded flight data used for flight operations safety improvement since all certified airlines require implementation of a safety and quality management system. The safety performance of a flight has been a challenging issue in the aviation industry and plays an important role to acquire competitive benefits. In this study, an integrated multi-class classification machine learning models and Markov chain were developed for cockpit crew performance evaluation during their flights. At the outset, the main features related to a flight are identified based on the literature review, flight operations expert’s statements, and the case study dataset (as numerical example). Afterwards, the flights’ performance is evaluated as a target column based on four multi-class classification models (Decision Tree, Support Vector Machine, Neural Network, and Random Forest). The results showed that the random forest classifier has the greatest value in all evaluation metrics (i.e., accuracy = 0.90, precision = 0.91, recall = 0.97, and F1-score = 0.93). Therefore, this model can be used by the airline companies to predict flight crew performance before the flight in order to prevent or decrease flight safety risks.
期刊介绍:
CONCERNING THE FOLLOWING FIELDS OF RESEARCH: ▪ Flight Physics ▪ Air Traffic Management ▪ Aerostructures ▪ Airports ▪ Propulsion ▪ Human Factors ▪ Aircraft Avionics, Systems and Equipment ▪ Air Transport Technologies and Development ▪ Flight Mechanics ▪ History of Aviation ▪ Integrated Design and Validation (method and tools) Besides, it publishes: short reports and notes, reviews, reports about conferences and workshops