
Research Article
Predicting Vietnamese Stock Market Using the Variants of LSTM Architecture
@INPROCEEDINGS{10.1007/978-3-030-92942-8_11, author={Cong-Doan Truong and Duc-Quynh Tran and Van-Dinh Nguyen and Huu-Tam Tran and Tien-Duy Hoang}, title={Predicting Vietnamese Stock Market Using the Variants of LSTM Architecture}, proceedings={Nature of Computation and Communication. 7th EAI International Conference, ICTCC 2021, Virtual Event, October 28--29, 2021, Proceedings}, proceedings_a={ICTCC}, year={2022}, month={1}, keywords={LSTM Bidirectional LSTM VN-INDEX Stock market prediction}, doi={10.1007/978-3-030-92942-8_11} }
- Cong-Doan Truong
Duc-Quynh Tran
Van-Dinh Nguyen
Huu-Tam Tran
Tien-Duy Hoang
Year: 2022
Predicting Vietnamese Stock Market Using the Variants of LSTM Architecture
ICTCC
Springer
DOI: 10.1007/978-3-030-92942-8_11
Abstract
Recently, the problem of stock market prediction has attracted a lot of attention. Many studies have been proposed to apply to the problem of stock market prediction. However, achieving good results in prediction is still a challenge in research and there are very few studies applied to Vietnamese stock market data. Therefore, it is necessary to improve or introduce new forms of prediction. Specifically, we have focused on the stock prediction problem for the Vietnamese market in the short and long term. Long short-term memory (LSTM) based on deep learning model has been applied to big data problem such as VN-INDEX. We compared the prediction results of the variants of the LSTM model with each other. The results obtained are very interesting that the Bidirectional LSTM architecture gives good results in short- and long-term prediction for the Vietnamese stock market. In conclusion, the LSTM architecture is very suitable for the stock prediction problem in the long- and short- term.