摘要
将上证指数和深证成指股票数据作为研究对象,对股票长期价格进行预测.选取长短期神经网络、循环神经网络、HP滤波长短期神经网络混合模型和HP滤波循环神经网络混合模型进行比较分析.经过模型间的对比分析,发现HP滤波对长短期神经网络预测的优化效果要优于循环神经网络.
Using the Shanghai composite index and Shenzhen composite index stock data as research objects,the long-term stock prices is predicted.Long and short term neural network,recurrent neural network,HP filtered long and short term neural network hybrid models and HP filtered recurrent neural network hybrid models are selected for comparative analysis.Through comparative analysis between models,it is found that HP filtering has better optimization effects on the long and short term neural network models than recurrent neural network models.
作者
郑爱宇
孙德山
ZHENG Aiyu;SUN Deshan(School of Mathematics,Liaoning Normal University,Dalian 116029,China)
出处
《高师理科学刊》
2023年第10期36-40,共5页
Journal of Science of Teachers'College and University
基金
辽宁省教育厅项目资助(LJKMZ20221424)。
关键词
长短期神经网络
循环神经网络
股票预测
HP滤波
long and short term neural network model
recurrent neural network model
stock preparation
HP filter