BP神经网络在股票预测中的应用研究

The Research in Stock Forecast Based on BP Neural Network

分析研究了BP算法,并提出了一些相应的改进措施。将改进的BP算法应用于股票数据分析,即利用BP网络很强的非线性映射功能,模拟影响股票的相关因素与股票开盘价、收盘价之间的关系,建立了基于BP神经网络的股票预测系统。用2004年的沪市证券数据进行了验证,结果表明此方法不仅收敛速度快、预测精度高,而且具有一定应用价值。

This paper analysis the BP algorithm, proposes some improved methods and make use of the BP network＇s strong functions to establish a stock forecast system. Finally, the testing results indicate that not only the convergent rate is quick, but also the forecast is precision.