基于Elamn神经网络的股票价格预测研究.PDFVIP

第 26卷第 2期 计算机应用 Vol. 26 No. 2 2006年 2月 Computer pplications Feb. 2006 : 1001- 9081( 2006) 02- 0476- 02 Elman 1, 2 2 林春燕 , 朱东华 ( 1. 北京理工大学 应用数学系, 北京 100081; 2. 北京理工大学 管理与经济学院, 北京 100081) ( lchy26@ sohu. com) :为了更好地把握股票价 的波动,应用了在处理序列数据输入输出具有优越性的 E lman 递归神经网络建立股市预测模型, 并用两支股票进行了检测, 检测结果说明人工神经网络应用于中国 股票市场的预测是可行和有效的, 有着良好的前景 : Elman神经网络; 预测;股票 : TP311 : Stock m arket forecasting research based on E lm an neural n etw ork 1, 2 2 LIN Chun-yan , ZHU Dong-hua (1. D epartmen t of M athem atics, B eij ing Institute of Technology, B eij ing 100081, Ch ina; 2. S chool of M anagemen t E conom ics, Beij ing Institu te of T echnology, Beij ing 100081, China ) A bstract: The Stock market is typical complex object, which is hard to construct accurate mathematical model. For holding the securities price volatility, to presents amethod for stockmarket prediction modeling by utilizing the good handles date capability ofElman neuralnetworks. nd the modelwas tested by wt o stocks. The result of the tests show that Elman neural newt orks are feasible and efficient for the prediction of Chinese stockmarket. K ey w ords: Elman neural network; prediction; Stock 0 引言 1 Elman股价预测模型网络结构 1. 1 Elman , , , Elman Elman 1990 , Elman , , R IM G RCH , Elman, 1, , J. H. W ang J. y. Leu 1996 , BP RIM , Sung-suk Kim 1998 , ,