Automobile Traffic Accidents Prediction Model Using by Artificial Neural Networks

Abstract

As increasing environmental factors in automobile traffic accidents, combination works of researches such as driving status and information analysis are also increasing. To prevent accidents, these factors are to be eliminated finally but it could not best solution due to time and space limitation. The data mining technique is also applied in various fields as a method to extract information based on massive data, and neural networks are also utilized as useful modeling technique. In this paper, using the neural network in a traffic accident in-depth analysis of scientific research through the pre-crash factors is proposed in order to reduce traffic accidents. For the prevention of traffic accident, the main factors are found associated with deaths. However, it is a difficult problem to present the perfect result while considering all circumstances if applied in actual problems, as multiple variables that affects the result. Through neural network learning values by adjusting the weights between each node and important factors can be found. In this paper, two kinds of neural networks of MLP (Multi Layer Perceptron) and RBFN (Radial-Basis Function Network) are experimented by XLMiner. As the result, it had some weight whether driving with drunk or high speed.