非线性代数模型LSTM的应用研究——以股票指数预测为例  

Research on the Application of LSTM for Nonlinear Algebraic Models--Take Stock Index Forecasts as an example

作　　者：郭华毅 GUO Hua-yi(Shanxi Pharmaceutical Vocational College,Taiyuan 030031,China)

机构地区：[1]山西药科职业学院,山西太原030031

出　　处：《江西电力职业技术学院学报》2023年第1期139-141,144,共4页Journal of Jiangxi Vocational and Technical College of Electricity

摘　　要：为合理预测我国股票指数,建立了一个深层LSTM非线性模型。根据新浪财经提供的2020年1月1日至12月31日,中国股票市场18个指数相关的数据,运用LSTM非线性模型预测该18个指数的收盘价,并将LSTM非线性模型的预测性能与ARIMA线性模型和SVR非线性模型进行对比。实验表明,在短期与长期预测性能对比中,LSTM非线性模型的预测性能最佳。In order to reasonably predict China’s stock index,a deep LSTM nonlinear model is established.According to the data related to 18 indices in China’s stock market from January 1,2020 to December 31,2020 provided by Sina Finance,the LSTM nonlinear model was used to predict the closing prices of the 18 indexes,and the prediction performance of the LSTM nonlinear model was compared with the ARIMA linear model and SVR nonlinear model.Experiments show that the LSTM nonlinear model has the best prediction performance in the comparison of short-term and long-term prediction performance.

关 键 词：股票指数 预测 LSTM非线性模型 ARIMA线性模型 SVR非线性模型 预测性能 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]