Flight accident modeling and predicting based on least squares support vector machine

Abstract

To analyze and predict the flight safety situation scientifically, a flight accident prediction model, based on least squares support vector machine (LS-SVM), is proposed. Simultaneously, three steps grid search method is introduced into the flight accident modeling to select efficiently the parameters of LS-SVM. The simulation results of modeling and predicting on flight accident 100000-hour-rate indicate that, LS-SVM model and its parameter selection method are feasible and effective in flight accident prediction.