基于时间序列分析的股价预测  

Stock-price Forecasting Based on time Series Analysis

作　　者：方启东[1] 温鑫[1] 蒋佳静[1] 丁攀攀[1] 沈友红[1] 王琰[1] 

机构地区：[1]宿州学院应用数学系,安徽宿州234000

出　　处：《宿州学院学报》2010年第8期71-73,F0003,共4页Journal of Suzhou University

基　　金：宿州学院大学生科研课题项目(KYLXLKY-200915)

摘　　要：基于收盘深发展股票价格的实际数据资料,通过对实际样本数据的预处理,确认股票价格序列为平稳非白噪声序列的基础上,利用SAS/ETS软件,采用ARMA方法,建立时间序列预测模型。应用的结果表明:本预测模型考虑了各种随机因素的影响,与实际状况有较高的吻合度,对动态数据进行预测是可行的,对于经济统计预测有一定的实用价值。Based on the practical datum of stock price about Shenfazhan（a company）,the datum are pre-processed and the stock price is considered to be steady non-white noise series.The time series predictive model is set up by SAS/ETS software and also ARMA method is used.ARMA method is a predictive method with great accuracy in short-term time series.The result illustrates that this forecast model has considered each kind of random factors influence,and it has the high fitness with the actual condition.It proves the practicability and effectiveness of this predictive model in forecasting on the basis of dynamic datum.The application of time series forecasting method in this essay has a certain practical value in the statistics forecast in the field of economics.

关 键 词：平稳时间序列 股价 预测 ARMA模型 

分 类 号：F830.91[经济管理—金融学]