基于股票预测的ARIMA模型、LSTM模型比较  

Comparison of ARIMA Model and LSTM Model Based on Stock Forecast

作　　者：丁文绢 

机构地区：[1]西华大学经济学院,四川成都610039

出　　处：《工业控制计算机》2021年第7期109-112,116,共5页Industrial Control Computer

摘　　要：股票市场是金融市场的重要组成部分,与经济的发展密切相关。对于股票价格的各种分析预测问题伴随着金融市场的建立一直存在,为此使用上证A股50的历史交易数据作为研究对象,对其进行收盘价格趋势预测分析。通过ARIMA模型、LSTM模型对股价走势进行预测。经过实证研究,结合误差指标和交易绩效等展示模型预测精度和预测效果,最后得出基于LSTM模型的深度神经网络模型具有较好的预测精度。并且通过使用多种深度学习方法,从金融市场的历史交易数据中发现当前市场中潜在的获利机会,指导机构和个人投资者进行更好的投资。This paper uses the historical transaction data of the Shanghai A-share 50 as the research object to carry out forecasting and analysis of the closing price trend. Predict the stock price trend through ARIMA model and LSTM model.After empirical research,combined with error indicators and transaction performance to show the model’s forecasting accuracy and forecasting effect,it is finally concluded that the deep neural network model based on the LSTM model has better forecasting accuracy.And by using a variety of deep learning methods,we can discover potential profit opportunities in the current market from historical transaction data in the financial market,and guide institutions and individual investors to make better investment behaviors.

关 键 词：深度学习 Seq2seq 神经网络 价格趋势预测 

分 类 号：F832.51[经济管理—金融学]