基于改进的Elman神经网络的股价预测模型

Elman神经网络是一种典型的回归神经网络,比前向神经网络具有更强的计算能力,具有适应时变特性的能力,因而非常适用于对股市这一类极其复杂的非线性动力学系统进行预测。文中以深市A股中的个股中集集团(股票代号:000039)的共180天的实际收盘价的时间序列作为预测对象,提出基于改进的Elman神经网络的个股价格预测模型,实验结果取得较高的预测精度、较为稳定的预测效果和较快的收敛速度。这表明该预测模型对于个股价格的短期预测是可行和有效的。

Stock Price Forecasting Model Based on Improved Elman Neural Network

Elman neural network is a classical kind of recurrent neural network.It has greater calculating capability than feed forward neural network.And its characteristic of adapting time variability especially adapts to complicated nonlinear dynamics system forecasting like stock market.Uses the market price of Zhongji company(No.000039) in Shenzhen stock market A index as forecasting object and improved Elman neural network as forecasting model.The experiment results get higher precision,steadier forecasting effect and more rapid convergence speed.These show that model is feasible and efficient to forecast short term stock price.