Research of RNN Models Performance on New York Stock

DOI

10.2991/assehr.k.220401.121How to use a DOI?

Keywords

Artificial intelligence; Recurrent neural networks; Stock market prediction

Abstract

Stock market forecasts have become a popular topic for researchers and investors. Stock market forecasting methods range from traditional analysis based on statistics to machine learning models such as decision trees, SVM, and neural networks. In this project, we decided to use Recurrent Neural Network (RNN) models to make stock price forecasts due to the time series nature of stock prices. From simple RNN models to more complex models such as the GRU and LSTM, three different RNN models have been used to compare error values and the performance of each. Based on the results, we found that the LSTM was taking a longer time to train but better performance compared to the other two simple models. This RNN stock forecast study lays the foundation for the future use of RNN models in economic markets.

Copyright

© 2022 The Authors. Published by Atlantis Press SARL.

Open Access

This is an open access article distributed under the CC BY-NC 4.0 license.

Download article (PDF)

TY  - CONF
AU  - Zhentong Fan
AU  - Yang Wang
AU  - Cheng Ma
PY  - 2022
DA  - 2022/04/08
TI  - Research of RNN Models Performance on New York Stock
BT  - Proceedings of the 2022 International Conference on Social Sciences and Humanities and Arts (SSHA 2022)
PB  - Atlantis Press
SP  - 637
EP  - 642
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.220401.121
DO  - 10.2991/assehr.k.220401.121
ID  - Fan2022
ER  -