Abstract:

        This paper predicts the number of civil aviation unsafe event factors and stages with SVM and RBF neural network weighted combination model to improve the level of aviation safety and predict major flight accident risk, which is based on the accident data from Feb.2002 to Feb.2014. Its feasibility and effectiveness are proved by the comparison for the root mean square errors of three typical prediction methods. This can provide a scientific reference for the administration of civil aviation security.

Key words: prediction of unsafe events, support vector machine, RBF neural network