基于关联神经网络的股票预测模型  

Study on Stock Prediction Model Based on Correlative Neural Network

作　　者：陈阳[1] 宋玉坤[1] 

机构地区：[1]承德石油高等专科学校数理系,河北承德067000

出　　处：《承德石油高等专科学校学报》2008年第1期38-40,43,共4页Journal of Chengde Petroleum College

摘　　要：通过关联分析从众多的技术指标中找到与日收盘价关联度较大的技术指标作为输入变量,利用BP网络较好的分类能力,结合国内股票市场的特性,对于个股中国首钢日收盘价涨跌的短期预测进行了初步的探讨。数值实验结果表明关联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。Through correlation analysis, the author puts technical indexes as input variables from technical indicators closely associated with the closing price. With the help of BP network＇s better classification and the characteristics of domestic stock market, the author discusses the short-term forecasts of the closing price pullback of Shougang stock. Numerical experiments show that the neural network used in forecasting China＇ s stock market is feasible and effective, and with a prospective future.

关 键 词：股票 BP神经网络 关联分析 技术指标 

分 类 号：F830.91[经济管理—金融学]