基于GRU的股票涨跌预测  

GRU-based Stock Ups and Downs Forecast

作　　者：赵建群[1] 王悦[1] ZHAO Jianqun;WANG Yue(School of Economics and Management, Guangdong Polytechnic of Science and Trade, Guangzhou 510000, China)

机构地区：[1]广东科贸职业学院经济管理学院,广东广州510000

出　　处：《湖州师范学院学报》2020年第8期93-98,共6页Journal of Huzhou University

摘　　要：为预测股票价格的涨跌,提出一种基于门控循环网络(GRU)方法.通过对涨跌幅度做多值量化,并根据其涨跌幅分布将涨跌进行3分类,模型的输入是中证100成分股的股票基本交易特征,利用GRU对其训练,最后对股票的涨跌做分类预测.实验结果表明,该模型在预测中证100整体股票价格涨跌上有较好的预测效果;对某一只股票的预测准确率与选取的股票有关.In order to predict the rise and fall of stock prices,a method based on gated cyclic network(GRU)is proposed.This paper quantifies the range of rise and fall and classifies the rise and fall into 3 categories according to the distribution of rise and fall.The input of the model is the basic trading characteristics of the 100 constituent stocks of the China securities regulatory system.The experimental results show that the model has a good effect in predicting the rise and fall of the overall stock price of CSI100.At the same time,in the prediction of a certain stock,the prediction accuracy is related to the selected stock.

关 键 词：中证100 GRU 涨跌预测 

分 类 号：F832[经济管理—金融学]