摘要
准确预测私人汽车拥有量,对制定经济政策和进行经济宏观调控、保证社会经济和谐发展有重要的作用。基因表达式编程(GEP)是新的进化模型,在数据挖掘领域得到了广泛的关注和研究,对符号回归任务表现了很强的优势。阐述了GEP基本原理,GEP进行序列分析的基本方法;根据1990—2007年全国和人汽车拥有量,基于GEP技术挖掘到了其模型。实验表明,基于GEP技术得到的私人汽车拥有量模型预测精度高、泛化能力强。
For making economic policies and controlling macro economy, it needs predict the total numbers of private cars. It is a vital role to ensure harmoniously develops the economics. Gene expression programming （GEP） is a novel evolution system and attracts many studies and attention. This paper introduced the principle of GEP, and the basic methods applied GEP to time series analysis. According to the total number of private cars from 1990 to 2007, based on GEP techniques, mined and analyzed prediction model. Illustrating from experiments, the model has a high prediction precision, and good generalization ability.
出处
《计算机应用研究》
CSCD
北大核心
2010年第3期958-960,共3页
Application Research of Computers
基金
江苏技术师范学院博士启动基金资助项目(KYY09001)
关键词
基因表达式编程
私人汽车量
预测
泛化能力
gene expression programming（GEP）
total number of private cars
prediction
generalization ability
