深度学习算法在股票价格预测方面的研究  

Research on Deep Learning Algorithm in Stock Price Prediction

作　　者：齐晓娜 张宇敬 郑艳娟[2] 封二英 QI Xiaona;ZHANG Yujing;ZHENG Yanjuan;FENG Erying(Hebei University of Finance Information Engineering and Computer School,Baoding Hebei 071000;student office of Hebei University of Finance,Baoding Hebei 071000)

机构地区：[1]河北金融学院信息工程与计算机学院,河北保定071000 [2]河北金融学院学生处,河北保定071000

出　　处：《软件》2021年第7期70-72,共3页Software

基　　金：河北省教育厅科技重点项目(ZD2019136)。

摘　　要：对于金融机构或个人投资者而言,对股票价格进行精准预测,进而制定优化的量化投资策略,可以在降低投资风险的同时实现利益最大化。基于传统的金融模型或是传统的机器学习模型都不能达到理想效果。本文采用深度学习算法,建立了适用于股票预测的长短记忆网络(LSTM)模型,通过该模型分别对中国人寿和长城汽车两只股票数据进行训练验证,实验表明该模型可以获得比较优异的预测精度。For financial institutions or individual investors,accurate prediction of stock prices and formulationof optimized quantitative investment strategies can reduce investment risks and maximize benefits.Neither thetraditional financial model nor the traditional machine learning model can achieve the ideal effect.In this paper,Long short-term memory(LSTM)model suitable for stock forecasting is established,and the model is trained andverified by the stock data of China Life(601628)and GreatWall(601633).The experiments show that the model canobtain excellent prediction accuracy.

关 键 词：股票预测 深度学习 LSTM 

分 类 号：F832.48[经济管理—金融学]