基于非参数回归的交通事件影响范围预测

Calculate the Affected Area of Traffic Incident Base on Non-Parametric Regression

基于非参数回归方法,创新性地利用拥堵信息建立模式库,解决了传统分类、回归、神经网络等预测方法的事件特征选取限制及训练样本不足等问题,提高了预测结果的准确性。

This paper proposes the prediction model based on nonparametric regression through establishing Model Base by traffic con- gestion information. Then the model solves the problems of limit in feature select and lack in training samples among traditional classi- fication, regression analysis and Neural Network. Finally improve the accuracy of prediction result effectively.