支持向量机在出行链模式识别和影响因素分析中的应用（英文）

发布时间：2018-04-15 08:36

本文选题：出行链模式 + 支持向量机　；参考：《Journal of Southeast University(English Edition)》2017年01期

【摘要】：为了提高交通需求预测精度,综合考虑居民出行行为在时间维度上的分布,采用支持向量机、径向基神经网络和多项logit三种方法,基于居民活动目的,建立了出行链模式识别模型,并利用敏感性分析方法研究了解释因素对出行链模式选择的影响和对模型性能的贡献程度.结果显示:支持向量机模型在总体准确度和分类准确度上均优于其他2种方法,体现了支持向量机在小样本下的识别性能优势;证明了支持向量机能够较准确地反映多分类因素对于出行链模式选择行为的影响程度;因素对于不同出行链模式识别精度的贡献度差异表明了细化出行链模式及探索各个模式特有影响因素的重要性.支持向量机技术在交通需求预测建模及影响因素分析方面均具有实践意义.
[Abstract]:In order to improve the accuracy of traffic demand prediction, considering the distribution of travel behavior in time dimension, support vector machine (SVM), radial basis function neural network (RBF) and multi-item logit are adopted, which are based on the purpose of residents' activities.A trip chain pattern recognition model is established, and the influence of explanatory factors on trip chain pattern selection and its contribution to model performance are studied by sensitivity analysis.The results show that the support vector machine model is superior to the other two methods in the total accuracy and classification accuracy, which shows the superiority of support vector machine in small sample recognition.It is proved that support vector machine can accurately reflect the influence of multi-classification factors on the behavior of trip chain pattern selection.The difference in the contribution of factors to the recognition accuracy of different trip chain patterns indicates the importance of refining trip chain patterns and exploring the unique influencing factors of each pattern.Support vector machine (SVM) technology has practical significance in traffic demand forecasting modeling and factor analysis.
【作者单位】：东南大学交通学院;
【基金】：The Fundamental Research Funds for the Central Universities,the Scientific Innovation Research of College Graduates in Jiangsu Province(No.KYLX_0177)
【分类号】：U491

【相似文献】