EUVIMEDEUVIMED entdecken
Crossref

A novel LASSO-ATT-LSTM model of stock price prediction based on multi-source heterogeneous data

Huiru Li, Yanrong Hu, Hongjiu Liu

Journal of Intelligent & Fuzzy Systems · 01.06.2023 · 44 (6) · 10511-10521

DOI: 10.3233/jifs-221919 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent & Fuzzy Systems, erschienen am 01.06.2023. Verfasst von Huiru Li und 2 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 37 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Stock price volatility is influenced by many factors, including unstructured data that is not easy to quantify, such as investor sentiment. Therefore, given the difficulty of quantifying investor sentiment and the complexity of stock price, the paper proposes a novel LASSO-ATT-LSTM intelligent stock price prediction system based on multi-source data. Firstly, establish a sentiment dictionary in the financial field, conduct sentiment analysis on news information and comments according to the dictionary, calculate sentiment scores, and then obtain daily investor sentiment. Secondly, the LASSO (Least absolute shrinkage and selection operator) is used to reduce the dimension of basic trading indicators, valuation indicators, and technical indicators. The processed indicators and investor sentiment are used as the input of the prediction model. Finally, the LSTM (Long short-term memory) model that introduces the attention mechanism is used for intelligent prediction. The results show that the prediction of the proposed model is close to the real stock price, MAPE, RMSE, MAE and R2 are 0.0118, 0.0685, 0.0515 and 0.8460, respectively. Compared with the existing models, LASSO-ATT-LSTM has higher accuracy and is an effective method for stock price prediction.

Zitieren

APA

Li, H., Hu, Y., & Liu, H. (2023). A novel LASSO-ATT-LSTM model of stock price prediction based on multi-source heterogeneous data. Journal of Intelligent & Fuzzy Systems, 44(6), 10511-10521. https://doi.org/10.3233/jifs-221919

Vancouver

Li H, Hu Y, Liu H. A novel LASSO-ATT-LSTM model of stock price prediction based on multi-source heterogeneous data. Journal of Intelligent & Fuzzy Systems. 2023;44(6):10511-10521. doi:10.3233/jifs-221919

BibTeX
@article{Li2023,
  author = {Li, Huiru and Hu, Yanrong and Liu, Hongjiu},
  title = {A novel LASSO-ATT-LSTM model of stock price prediction based on multi-source heterogeneous data},
  journal = {Journal of Intelligent & Fuzzy Systems},
  year = {2023},
  volume = {44},
  number = {6},
  pages = {10511--10521},
  doi = {10.3233/jifs-221919}
}
Kostenlos starten →