Copula EDA-BP混合优化算法预测股票价格

股票价格的预测是广大投资者非常关注的问题,也是诸多学者不断研究的方向,神经网络具有学习样本规律的特点,通过神经网络预测股票价格是近几年研究的重点之一。Copula EDA-BP混合优化算法是利用了copula EDA的全局寻优和BP算法局部求精的特点,将两者结合起来建立了基于copula EDA-BP的模型系统,优化神经网络的权值阈值,对股票上证180的收盘价进行预测得到误差率,结果显示copula EDA-BP算法平均误差率低于BP算法,提高了传统BP神经网络的计算精度。

Copula EDA-BP Hybrid Optimization Algorithm to Predict the Stock Price

Stock price prediction is the issue thAt A vAst number of investors Are concerned And mAny scholArs Are continuing to study the direction. NeurAl network hAs the chArActeristics of leArning sAmples of neurAl network prediction,And the price of the stock is one of the reseArch focuses in recent yeArs. CopulA EDA-BP hybrid optimizA-tion Algorithm uses the globAl copulA EDA optimizAtion And BP Algorithm for locAl refinement to combine the two estAblished model system bAsed on copulA EDA-BP,weight And threshold optimizAtion neurAl network,predicted er-ror rAte on the Stock ExchAnge 180 closing price. Results show thAt the AverAge error rAte of the copulA EDA-BP algorithm is less thAn thAt of BP Algorithm,which improves the AccurAcy of trAditionAl BP neurAl network.