基于线性回归模型的铁路客运量预测与实证分析
Prediction and Empirical Analysis of Railway Passenger Volume Based on Linear Regression Model
王家俊1
作者信息
- 1. 中铁第五勘察设计院集团有限公司 郑州分院,河南 郑州 450001
- 折叠
摘要
以郑州市为例,利用计量经济学的方法对郑州市铁路客运量的影响因素进行单变量和多变量分析,采用普遍最小二乘法(OLS),分别研究地区生产总值、总人口数、接待国内外游客总量、居民消费价格指数对郑州市铁路客运量的影响程度.利用EViews软件,以2000-2016年的铁路客运量数据作为训练集,2017-2018年的铁路客运量数据作为测试集,分别进行一元线性回归和多元线性回归分析,通过多重共线性、异方差、自相关性的检验以及模型拟合优度的比较,得到最佳郑州市铁路客运量预测模型,模型解释能力为99.07%.经测试集相对误差检验,构建的模型预测值与实际值误差小于2%,多元线性回归模型预测效果很好,可为地区铁路客运量的短期预测及相关部门的客运决策提供数据支持.
Abstract
Taking Zhengzhou as an example,this paper makes univariate and multivariate analyses of the influencing factors of railway passenger volume in Zhengzhou City by the econometric method.It adopts the ordinary least square method(OLS)to study the influence degree of regional GDP,total population,total number of domestic and foreign tourists,and consumer price index on railway passenger volume in Zhengzhou City.Using EViews software,taking the railway passenger volume data from 2000 to 2016 as the training set and the railway passenger volume data from 2017 to 2018 as the test set,simple linear regression and multiple linear regression analysis were carried out respectively.Through multicollinearity,heteroscedasticity,autocorrelation test and model goodness of fit comparison,the best railway passenger volume prediction model in Zhengzhou was obtained,and the explanatory power of the model was 99.07%.The error between the predicted value and the actual value of the model is less than 2%through the relative error check of the test set.The multiple linear regression model has a good prediction effect,which can provide data support for the short-term prediction of regional railway passenger volume and passenger transport decision-making by relevant departments.
关键词
铁路客运量/线性回归模型/EViews软件/运量预测/模型检验Key words
railway passenger volume/linear regression model/EViews software/prediction of traffic volume/model check引用本文复制引用
基金项目
中铁第五勘察设计院集团有限公司科研计划(T5Y2020-C13)
出版年
2024
NETL
NSTL
万方数据