ESG Investment Analysis and Stock Price Prediction Using FinBERT and LSTM Models
Academic Conferences Series · 2025-12 · 18 (1)
DOI: 10.62381/acs.dimi2025.01 ↗Kurz gefasst
Fachartikel in der Zeitschrift Academic Conferences Series, 2025-12. Verfasst von Nan Zheng.
Abstract (Originalsprache)
This study explores the integration of Environmental, Social, and Governance (ESG) data with deep learning models for stock price prediction. We propose a hybrid framework combining FinBERT, a pre-trained financial language model, for sentiment analysis of ESG-related texts, and Long Short-Term Memory (LSTM) networks for time-series forecasting. Experimental results on Ping An Bank (stock code: 000001.SZ) demonstrate that the model achieves a Mean Absolute Error (MAE) of 0.107 and Root Mean Squared Error (RMSE) of 0.163, indicating robust convergence. The analysis highlights the potential of ESG sentiment as a predictive factor and underscores challenges such as data scarcity and inconsistent ESG ratings. Future research directions include multi-enterprise validation and model comparison.
Zitieren
APA
Zheng, N. (2025). ESG Investment Analysis and Stock Price Prediction Using FinBERT and LSTM Models. Academic Conferences Series, 18(1). https://doi.org/10.62381/acs.dimi2025.01
Vancouver
Zheng N. ESG Investment Analysis and Stock Price Prediction Using FinBERT and LSTM Models. Academic Conferences Series. 2025;18(1). doi:10.62381/acs.dimi2025.01
BibTeX
@article{Zheng2025,
author = {Zheng, Nan},
title = {ESG Investment Analysis and Stock Price Prediction Using FinBERT and LSTM Models},
journal = {Academic Conferences Series},
year = {2025},
volume = {18},
number = {1},
doi = {10.62381/acs.dimi2025.01}
}
EUVIMED