基于灰色Elman神经网络的股价预测

Stock price forecast based on grey Elman neural network

在灰色GM(1,1)模型动态预测股价变动趋势的基础上,运用Elman模型对其进行修正,采用组合模型的方法解决了非线性和影响因素复杂的股价预测的问题,并实证分析证明组合模型的可行性。预测结果表明:组合测试在保证稳定性的前提下提高了预测的准确率,更能拟合出股价变动的规律,为广大的投资者起到指导作用。

On the basis of dynamic prediction of stock price trend by grey GM(1,1)model,Elman model is used to revise it.The problem of stock price prediction with non-linearity and complicated influencing factors is solved by using combination model,and the feasibility of combination model is proved by empirical analysis.The forecasting results show that the combination test improves the forecasting accuracy on the premise of guaranteeing the stability,and can better fit the law of stock price fluctuation,which plays a guiding role for the vast number of investors.