基于因子分析和Logistic模型的中国客运量预测

发布时间：2018-05-18 08:28

本文选题：因子分析 + logistic模型　；参考：《系统管理学报》2014年03期

【摘要】：通过SPSS软件对1994~2011年中国客运量及影响因素进行因子分析,建立了中国客运量为被解释变量,其他影响因素为解释变量的多元统计分析模型。通过因子分析法分析影响中国客运量的主要因素,并进行了降维处理,消除由于变量过多而导致的多重共线性影响,同时得到反映不同年份的综合经济发展值的因子得分。考虑到2003年中国由于"非典"而导致的客运量降低的异常值的出现,将综合经济发展值与时间的关系首先作为一个整体研究,然后分为2个时间段分别研究,分别求出不同的综合经济发展值,构建了基于Logistic模型的中国客运量预测模型,进一步预测未来中国客运量。预测结果表明,通过较长时间段得到的综合经济发展值预测的数据误差相对较小。
[Abstract]:Based on the factor analysis of the passenger volume and its influencing factors in China from 1994 to 2011, the multivariate statistical analysis model of the passenger volume in China is established, which is the explanatory variable and the other influencing factors are the explanatory variables. This paper analyzes the main factors that affect the passenger volume of China by factor analysis, and carries on the dimensionality reduction treatment to eliminate the multiple collinear influence caused by the excessive variable, and obtains the factor score which reflects the comprehensive economic development value of different years at the same time. Considering the appearance of abnormal value of passenger traffic volume decrease caused by SARS in China in 2003, the relationship between comprehensive economic development value and time is first studied as a whole, and then divided into two time periods respectively. The different comprehensive economic development values are calculated, and the forecast model of Chinese passenger traffic volume based on Logistic model is constructed to predict the future passenger traffic volume in China. The prediction results show that the data error of the comprehensive economic development value predicted by a long period of time is relatively small.
【作者单位】：南京航空航天大学经济与管理学院;烟台大学外国语学院;徐州工程学院经济学院;
【基金】：国家社会科学基金资助项目(71073079);国家社会科学基金青年项目(12CGL042,12BGL104) 教育部人文社科基金资助项目(10YJC630129) 山东省自然科学基金资助项目(ZR2011GL019)
【分类号】：U116

【参考文献】