To effectively predict traffic fatalities and promote the friendly development of transportation, a prediction model of traffic fatalities is established based on support vector machine (SVM). As the prediction accuracy of SVM largely depends on the selection of parameters, Particle Swarm Optimization (PSO) is introduced to find the optimal parameters. In this paper, small sample and nonlinear data are used to predict fatalities of traffic accident. Traffic accident statistics data of China from 1981 to 2012 are chosen as experimental data. The input variables for predicting accident are highway mileage, vehicle number and population size while the output variables are traffic fatality. To verify the validity of the proposed prediction method, the back-propagation neural network (BPNN) prediction model and SVM prediction model are also used to predict the traffic fatalities. The results show that compared with BPNN prediction model and SVM model, the prediction model of traffic fatalities based on PSO-SVM has higher prediction precision and smaller errors. The model can be more effective to forecast the traffic fatalities. And the method using particle swarm optimization algorithm for parameter optimization of SVM is feasible and effective. In addition, this method avoids overcomes the problem of “over learning” in neural network training progress


    Access

    Download


    Export, share and cite



    Title :

    Traffic fatalities prediction based on support vector machine


    Contributors:
    Ting Li (author) / Yunong Yang (author) / Yonghui Wang (author) / Chao Chen (author) / Jinbao Yao (author)


    Publication date :

    2016




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    Unknown




    Reducing Commercial Aviation Fatalities Using Support Vector Machines

    Mishra, Anupam / Shrivastava, Kamal Krishna / Anto, Alex B. et al. | IEEE | 2019


    Fatalities of Heterogeneous Street Traffic

    Fazio, Joseph / Mazharul Hoque, M. / Tiwari, Geetam | Transportation Research Record | 1999


    Traffic Fatalities Early Assessment, 1984

    E. C. Cerrelli | NTIS | 1985


    Holiday Effect on Traffic Fatalities

    R. Arnold / E. C. Cerrelli | NTIS | 1987