Application Of Competitive Quantile Autoregressive Method In Aviation Accident Risk Estimation

With the steady growth of China’s national economy,the demand for air transportation will increase.China’s aviation infrastructure is gradually improving,and the proportion of air transportation in transportation is gradually increasing.Aviation accidents are causing social attention.Risk assessment and prevention of accidents is an important task for airlines.Through scientific assessment and prediction,airlines can make reasonable risk management decisions.Traditional time series models have rough data processing and cannot accurately predict risks.Therefore,there is an urgent need for improved models to be applied to aviation risk prediction.This article analyzes the data of aviation accidents in the United States.Firstly,the data characteristics are analyzed,and 15 variables including the number of casualties in accidents are selected to study their internal relationships using the FP Growth association rule algorithm.Next,the competitive quantile auto regression(CQAR)is used to estimate the aviation accident risk.Two example datasets are applied to CQAR,QAR model and ARMA model,verifying the asymptotic properties of CQAR and the excellence of the model,and also providing a more realistic scenario for the dynamic use of the model CQAR.The results show that the CQAR has the property of right asymptotic QAR,and for coverage,the CQAR algorithm using non optimal parameters is superior to the ARMA model.