基于潜在类别模型的城际高速铁路客运市场细分

Passenger market segmentation of inter-city high-speed railway based on latent class model

为了提高铁路企业经济收益,基于旅客的个人属性和行为特征研究城际高速铁路客运市场细分问题。在广珠城际高速铁路上进行问卷调查,采用潜在类别模型对旅客进行分类,选取性别、年龄、学历、税前月收入、出行平均次数、出行目的、出行费用作为外显变量,借助于Latent Gold 5.0软件对模型求解评价。研究结果表明:旅客被分为4类,相对误差均小于5%,且90%以上的旅客分类相对误差小于2.1%,说明了模型的合理性。前2类旅客的月收入和出行频率均较高,但在出行费用和目的方面差异较大,属于高端型旅客,建议为这些旅客设计高服务水平的运输产品。后2类旅客月收入和出行频率较低,但在年龄段和学历方面差异较大,属于低端型旅客,建议为这些旅客提供淡季折扣运输产品。

To improve the economic income of rail enterprises,the passenger market segmentation of the intercity high-speed railway was studied based on passengers’attributes and behavior features.In this paper,questionnaire survey was conducted on Guangzhou-Zhuhai railway and the latent class model was adopted to class passengers,taking the gender,age,education background,pre-tax monthly income,travel average times,travel purpose and travel expenses as the explicit variables.Latent Gold 5.0 was used to solve and evaluate the model and the results show that passengers are classified into 4 groups with the relative error less than 5%,and the relative error is less than 2.1%for more than 90%passengers,stating the model rationality.The first two groups feature with high monthly income and travel frequencies,but are greatly different in travel expenses and purposes. They belong to high-level passengers and train products with high-level services can be designed forthem. The other two groups feature with low monthly incomes and travel frequencies, but are greatly different inages and education backgrounds. They belong to low-level passengers and discount train products can bedesigned in slack seasons for them.