基于多元统计分析方法对我国民航客运量的研究
Research on Passenger Volume of Civil Aviation in China Based on Multivariate Statistical Analysis Method

本文以1978~1993年间我国民航客运量作为研究对象，选择“国民收入”，“消费总额”，“铁路客运量”，“民航航线里程”，“来中国旅行乘客数”5个指标。运用线性回归、主成分分析、聚类分析等多种多元统计分析方法，对影响我国民航客运量的主要因素进行分析，并探讨我国民航客运量与其它因素之间的具体函数关系。得到了以下结论：首先，假设检验的结果表明，铁路客运量、民航航线里程、来中国旅客数量是影响民航旅客运量的三个主要因素，其中民航航线里程与来中国旅行乘客数对我国民航客运量有显著的正向影响，而铁路客运量对我国民航客运量有显著的负向影响；另外，本文比较了由线性回归、岭回归与主成分回归三种方法拟合得到的回归方程，结果显示，岭回归法与主成分回归法所建立的回归模型对我国民航客运量的拟合程度更好，且更符合实际情况。
This paper took the passenger volume of China’s civil aviation during 1978~1993 as the research object, and selected five indicators: “national income”, “total consumption”, “railway passenger volume”, “civil aviation route mileage” and “the number of passengers traveling to China”. Using linear regression, principal component analysis, cluster analysis and other multi-variate statistical analysis methods, this paper analyzed the main factors affecting China’s civil aviation passenger volume and discussed the specific functional relationship between China’s civil aviation passenger volume and other factors. The following conclusions were obtained: First, the results of hypothesis test showed that the railway passenger volume, the civil aviation route mileage, and the number of passengers coming to China were the three main factors af-fecting the air route volume. The civil aviation route mileage and the number of passengers coming to China had a significant positive impact on the passenger volume of China’s civil avia-tion, while the railway passenger volume had a significant negative impact on passenger volume of China’s civil aviation. In addition, the regression equations fitted by linear regression, ridge regression and principal component regression were compared in this paper. The results showed that the regression model built by ridge regression and principal component regression fitted passenger volume of China’s civil aviation better and was more suitable for the practical situation.