论文部分内容阅读
提出了基于决策树的快速路交通事件持续时间预报方法。根据调查的快速路交通事件样本信息,提出了先通过聚类分析将连续型属性将交通事件持续时间离散化,再基于不纯性度量的决策树生成算法建立预测模型。研究分别采用方差分析方法对建模样本进行显著性检验和实际测试样本数据分别对模型进行了验证。结果证明所述的方法对于快速路交通事件预报是有效的。
A method of forecasting the duration of traffic incidents based on decision tree is proposed. According to the information of the samples of traffic accidents surveyed in the expressway, the paper proposes a method to establish the prediction model by discretely analyzing the duration of traffic incidents by using continuous attributes and the decision tree generation algorithm based on the impure measures. The study used variance analysis method to test the significance of the model sample and the actual test sample data respectively to validate the model. The results show that the proposed method is effective in predicting expressway traffic accidents.