Title

XGBoost and CNN-LSTM hybrid model with Attention-based stock prediction

Authors

Renzhe Zhu, Wenzhou-Kean University
Yingke Yang, Wenzhou-Kean University
Junqi Chen, Wenzhou-Kean University

Document Type

Conference Proceeding

Publication Date

1-1-2023

Abstract

The stock market is crucial to the growth of the economy. Research and predictions on the change in stock price might help investors minimize risk because of the stock market's complicated volatility. Traditional time series models, such as ARIMA, cannot adequately represent nonlinearity and produce accurate stock forecasts. According to this study, a hybrid model combining attention-based CNN-LSTM and XGBoost demonstrates the ability to predict stock prices effectively, leveraging the high nonlinear generalization capabilities of neural networks. The research introduces a novel approach that integrates various components, including a time series model, CNN with Attention mechanism, LSTM, and XGBoost regressor, in a nonlinear connection.

Publication Title

2023 IEEE 3rd International Conference on Electronic Technology, Communication and Information, ICETCI 2023

First Page Number

359

Last Page Number

365

DOI

10.1109/ICETCI57876.2023.10176988

Recommended Citation

Zhu, Renzhe; Yang, Yingke; and Chen, Junqi, "XGBoost and CNN-LSTM hybrid model with Attention-based stock prediction" (2023). Kean Publications. 326.
https://digitalcommons.kean.edu/keanpublications/326