摘要
对BP神经网络的算法和结构进行了介绍,并基于MATLAB的BP网络的工具箱,对浦发银行近一年交易日的数据进行了训练和测试,获得了一定的预测精度。最后还对波动期股价预测的难易,股价涨幅预测的难度以及输入变量对BP网络预测股价能力的影响等进行了讨论。基于实验结果,得出了基于BP神经网络的数学模型能一定程度上实现对股价趋势判断的结论。
The paper analyses the algorithm and structure of the BP neural network , and has a train and test on the stock data of Shanghai Pudong Development Bank over the past year, based on the BP neural network toolbox of MATLAB, and finally gets a result of prediction with a certain accuracy. The paper also discusses the difficulty for predicting in the wave period of the stock price , the difficulty for predicting on the percentage of the stock prices up and down, and how did the input variables of BP network impact the result of the prediction. The result of the experiments tells that we can predict the trend of stock＂ s price on a certain degree based on the BP neural network.
出处
《浙江金融》
北大核心
2011年第11期57-59,64,共4页
Zhejiang Finance
