which is of course challenging, or merging multi-
source heterogeneous stock information. Although it
is thought that similar results can be drawn, the author
did not evaluate the methodologies used in this study
using data from other nations or industries due to time
and space constraints. These can all be taken into
account in further work.
Subsequent investigations will primarily modify
the model's parameters in an effort to increase the
findings' accuracy. Future research work will also
conduct model stability analysis on the proposed
model to study whether the model is applicable to
other data sets estimation in other application fields,
such as gold price prediction and weather forecast.
4 CONCLUSIONS
To sum up, this study assesses and contrasts the
performance of many hybrid deep learning models,
concentrating on Guizhou Moutai stock. Results
demonstrate that advanced models, especially CNN-
BiLSTM-AM, outperform simpler models like MLP
and RNN on the precision of the predictions. Using
CNN for extraction of features, BiLSTM for temporal
dependency extraction, and Attention Mechanism for
emphasizing key information leads to superior
predictive performance. The CNN-BiLSTM-AM
model achieves the closest R
2
to 1 and the lowest
RMSE and MAPE, highlighting its effectiveness in
handling complex time series data. In the future,
studies should concentrate on increasing model
efficiency in order to lower computing expenses,
making it suitable for real-time trading environments.
Additionally, exploring the integration of external
factors like macroeconomic indicators could further
improve prediction accuracy. The study's significance
lies in providing a comprehensive comparison of
predictive models and offering insights into how
hybrid architectures can enhance stock price
forecasting.
REFERENCES
Box, G. E. P., Jenkins, G. M., 1970. Time Series Analysis:
Forecasting and Control. Holden-Day.
Haykin, S., 1999. Neural Networks: A Comprehensive
Foundation. Prentice Hall.
Hochreiter, S., Schmidhuber, J., 1997. Long Short-Term
Memory. Neural Comput., 9, 1735–1780.
Lu, W., Li, J., Li, Y., Sun, A., Wang, J., 2020. A cnn-lstm-
based model to forecast stock prices. Complexity, 2020,
1–10.
Luo, J., Zhang, X., 2022. Convolutional neural network
based on attention mechanism and bi-lstm for bearing
remaining life prediction. Applied Intelligence, 52(2),
1076-1091.
Qin, L., Yu, N., Zhao, D., 2018. Applying the convolutional
neural network deep learning technology to behavioral
recognition in intelligent video. Tehnički vjesnik, 25(2),
528-535.
Sun, Y., Sun, Q., Zhu, S., 2022. Prediction of Shanghai
stock index based on investor sentiment and cnn-lstm
model. J. Syst. Sci. Inf. Technol. Engl. Ed., 10, 620–632.
Treisman, A., Gelade, G., 1980. A feature-integration
theory of attention. Cogn Psychol 12(1), 97–146
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones,
L., Gomez, A. N., Kaiser, Ł., Polosukhin, I., 2017.
Attention is all you need. Advances in Neural
Information Processing Systems, 30, 5998-6008.
Wang, J., Wang, J., Sun, Q., 2021. Forecasting stock
market with optimized model by artificial rabbits
optimization algorithm. Expert Systems with
Applications, 163, 113481
Xiao, C., Xia, W., Jiang, J., 2020. Stock price forecast
based on combined model of ARI-MA-LS-SVM. Neural
Comput Appl., 32(10), 5379-5388.
Yu, P., Yan, X., 2020. Stock price prediction based on deep
neural networks. Neural Comput Appl., 32(6), 1609-
1628.
Zhang, J., Ye, L., Lai, Y., 2023. Stock Price Prediction
Using CNN-BiLSTM-Attention Model. Mathematics,
11(1985), 1-18.