汽车潜在消费者的数据挖掘  

Data Mining in Obtaining Potential Car Consumers

作　　者：王晓峰[1] 苗利峰[2] 李颜[2] 

机构地区：[1]河北广播电视大学高等职业技术学院,河北石家庄050081 [2]河北科技大学理学院,河北石家庄050018

出　　处：《数学的实践与认识》2013年第17期285-289,共5页Mathematics in Practice and Theory

基　　金：河北省高等学校自然科学自筹基金(Z2011139);河北省教育厅优秀青年基金(Y2012020);石家庄市科学技术研究与发展指导计划(125790375)

摘　　要：由于目前私家车购买比例的迅速上升,汽车经销商之间的竞争也越来越激烈,如何尽可能的获取汽车潜在消费者变得尤为重要.通过利用国家统计局的"私家车预购数据",用支持向量机算法建立了预购车数据分类模型,模型通过对未来家庭轿车潜在消费者和非潜在消费者进行分类,为确定消费者在未来三年内买车的可能性提供科学依据,同时建立了决策树模型,用于比较与分析.At present, the competition of auto dealers will be more and more fierce as the ratio of purchasing a private car rises rapidly. How to achieving potential consumers as much as possible becomes particularly important. Thus, based on the database that was established by National Bureau of Statistics of China, a classification model was constructed by Support Vector Machines（SVM）, which can provide a scientific basis for forecasting the possibility that a new consumer will buy a car in the coming three years by classifying potential consumers and non-potential consumers. Meanwhile, a decision tree model was built for analysis and comparison with SVM.

关 键 词：支持向量机 潜在消费者 分类模型 决策树 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]