期刊名称:IAENG International Journal of Computer Science
印刷版ISSN:1819-656X
电子版ISSN:1819-9224
出版年度:2020
卷号:47
期号:3
语种:English
出版社:IAENG - International Association of Engineers
摘要:The recurrent neural network is generally utilized in an assortment of areas, such as pattern recognition, natural language processing and computational learning. Time series prediction is one of the most challenging topics for many years due to its application in finance and decision making. This article centers chiefly on the methods and techniques of forecasting future prices of two stocks: IAM.PA and ORA.PA. An experimental investigation is grounded on two years of historical information. Furthermore, the statistics of the stocks and the predictions are made for 22 days in advance. The prediction performance compares three approaches of the recurrent neural network: Elman recurrent neural network in the first stage, Long Short-Term Memory recurrent neural network for the next phase, and Gated Recurrent Unit in the third phase. The article aims to accommodate a comparative analysis among these three models based on mean square error, time per step, memory and the number of hidden nodes required for excellent accuracy.