BP神经网络误差修正模型的S&P500指数预测  

S&P500 index forecast based on BP neural network error correction model

作　　者：周万珍[1] 阚景森 ZHOU Wanzhen;KAN Jingsen(School of Information Science and Engineering,Hebei University of Science and Technology,Shijiazhuang 050018,China)

机构地区：[1]河北科技大学信息科学与工程学院,石家庄050018

出　　处：《中国科技论文》2018年第14期1649-1653,共5页China Sciencepaper

摘　　要：为克服BP神经网络在预测模型构建过程中容易陷入"局部最优"以及隐含层数目等参数选择不当容易造成"过拟合"或"欠拟合"等问题,基于支持向量机(SVM)构建了一种BP神经网络误差修正模型。首先通过神经网络实现对S&P500指数的预测,然后通过支持向量机构建S&P500指数涨幅情况预测模型,基于神经网络与支持向量机的两种预测结果构造S&500指数预测误差修正模型,实现对BP神经网络预测误差的修正。实验结果表明,在本文数据集下所构建的修正模型预测准确率明显优于BP神经网络。It is easy to fall into local optimal,and result in over-fitting or under-fitting caused by the improper selection of hidden layer numbers or other parameters in the process of forecasting model built using BP neural network.To overcome these limitations,a BP neural network error correction model is built based on support vector machine.Firstly,the forecast of S&P500 index is realized by neural network.Then,the increasingly forecasting model of S&P500 index is built by support vector machine.Finally,the prediction error of BP neural network is corrected by the error correction model built by both prediction results of neural network and support vector machine.Experimental results show that the accuracy of this model is better than BP neural network obviously in our datasets.

关 键 词：BP神经网络 支持向量机 误差修正模型 S&P500指数预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]