摘要
提出一种基于BP神经网络的股票价格预测模型SPPM(Stock Price Prediction Model)。SPPM集成了多个神经网络,可预测未来若干天的股价走势。针对SPPM中的数据预处理、输出融合、神经网络隐藏层节点数选取等关键问题作了详细讨论。实验结果表明,SPPM具有一定的实际价值。
A BP neural network based stock price prediction model( SPPM) is put forward. SPPM integrates a number of neural networks, so that it can predict stock price trends in a few days. Some key problems of SPPM such as data preprocessing,output fusion,neural network hidden layer node number setting and so on are discussed in detail. Experiment results show that SPPM is of certain practical value.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第5期89-92,共4页
Computer Applications and Software
基金
河北省科技厅项目(09213575D
09213515D)