首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于基因表达式编程的股票指数时间序列分析
引用本文:廖勇,唐常杰,元昌安,陈安龙,段磊.基于基因表达式编程的股票指数时间序列分析[J].四川大学学报(自然科学版),2005,42(5):931-936.
作者姓名:廖勇  唐常杰  元昌安  陈安龙  段磊
作者单位:1. 四川大学计算机学院,成都,610064
2. 四川大学计算机学院,成都,610064;广西师范学院信息技术系,南宁,530001
基金项目:国家自然科学基金(60473071),973计划项目(2002CBlll504),博士点基金(20020610007)
摘    要:基因表达式编程(GEP)是遗传算法研究的新分支.针对股票对象的特点,提出了适应股票规律的GEP—STOCK模型,包括n时段—STOCK—GENE,STOCK—fitness以及STOCK-GEP算法,并以上海证券交易指数时间序列数据为对象做了实验.进行了误差和指数涨跌分析.实验结果表明GEP—STOCK模型预测精度较高,20d的平均绝对误差为11.08,平均相对误差为0.64%.从涨跌情况预测来看,模型对6d后指数的涨跌判断,正确率高于80%以上.

关 键 词:数据挖掘  基因表达式编程  时间序列  股票数据
文章编号:0490-6756(2005)05-0931-06
收稿时间:2004-12-30
修稿时间:2004-12-30

Time Series Prediction in Stock-Price Index Based on Gene Expression Programming
LIAO Yong,TANG Chang-jie,YUAN Chang-an,CHEN An-long,DUAN Lei.Time Series Prediction in Stock-Price Index Based on Gene Expression Programming[J].Journal of Sichuan University (Natural Science Edition),2005,42(5):931-936.
Authors:LIAO Yong  TANG Chang-jie  YUAN Chang-an  CHEN An-long  DUAN Lei
Abstract:The Gene Expression Programming(GEP) is a new branch of Genetic Algorithm(GA).Based on the features of stock objects,presents the GEP-STOCK model including the STOCK-GENE and the STOCK-fitness that appropriated to the special rules of stocks,and STOCK-GEP algorithm,gives experiments and analysis on the real stock-price index of Shanghai Stock Exchange.The results show that the precision that predicts by using stock-model is higher than traditional method.The average of absolute error in 20 d is 11.08,the average of relative error is 0.64%.Analyzing the rise-fall of the stock-price index according to the STOCK-GEP model,the correct rate of the rise-fall index after 6 d is up to 80%.
Keywords:data mining  gene expression programming  time series  Stock data
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号