基于PCA-Logistic回归的汽车保有量预测研究  

Prediction of Car Ownership Based on Principal Component Analysis and Logistic Regression

作　　者：张兰怡[1] 胡喜生[1] 陈清耀 邱荣祖[1] 

机构地区：[1]福建农林大学交通与土木工程学院,福建福州350002

出　　处：《重庆交通大学学报（自然科学版）》2017年第5期104-109,共6页Journal of Chongqing Jiaotong University(Natural Science)

基　　金：福建省社会科学规划项目青年基金项目(FJ2015C148);福建省教育厅科技项目(JB14005);福建农林大学高水平大学建设基金项目(113-612014018);福建农林大学青年基金项目(2013xjj25)

摘　　要：汽车保有量是一个相对复杂、非线性变化的数据总量,需要一种预测方法对汽车保有量进行快速、准确、合理的预测,预测结果可以作为城市经济可持续发展的重要依据。以福建省为例,选取2000—2014年间福建省总人口、人均GDP、第一产业生产总值比重、第二产业生产总值比重、第三产业生产总值比重、城镇居民人均可支配收入、农村居民人均纯收入、城市化水平等8个指标作为汽车保有量的主要影响因素进行分析。对8个指标进行主成分分析得到综合经济发展值的预测方程,采用Logistic回归模型进行预测并验证。结果显示:该方法预测精度高,能够为对汽车保有量进行较准确的估计,并为城市发展规划提供参考依据。Car ownership is the amount of data with relatively complex and nonlinear changes. There is a need for a rapid and accurate prediction method for fast, accurate and reasonable prediction of car ownership, whose prediction results can be used as an important basis for the sustainable development of the city economy. A case study of Fujian province was carried out. 8 indicators were selected as main influence factors of car ownership in Fujian province from 2000 to 2014, such as total population, per capita GDP, primary industry proportion of GDP, the second industry proportion of GDP, the third industry proportion of GDP, urban per capita disposable income, rural per capita net income and urbanization level. Through principal component analysis on 8 indicators, the prediction equation of comprehensive economic development value was obtained, which was predicted and verified by Logistic regression model. It is indicated that the prediction accuracy of the proposed method is high, which can predict the ear ownership exactly and provide reference basis for urban development planning.

关 键 词：交通工程 汽车保有量 主成分分析 LOGISTIC回归模型 预测 

分 类 号：U491.14[交通运输工程—交通运输规划与管理]