EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

A study on short-term vegetable price prediction based on the CNN-LSTM-Attention model

Weijia Bao, Wuzheng Su, Xin Zhao, Jiayu Zhuang

Discover Food · 04.06.2025 · 5 (1)

DOI: 10.1007/s44187-025-00426-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Discover Food, online erschienen am 04.06.2025. Verfasst von Weijia Bao und 3 weiteren Autor:innen. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 72 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: the Tianshan Talent Training Program, the Technology Innovation Project Fund of the Chinese Academy of Agriculture Sciences.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In view of the lack of consideration of uncertainty factors in vegetable price prediction research and that short-term vegetable price prediction is crucial for stabilizing the industry's development, this paper focuses on the Chinese vegetable market. From the standpoint of perfecting feature variables, we add the Baidu index with extreme weather as the keyword and the Chinese economic policy uncertainty index as the new feature variables, while analyzing whether using the CNN-LSTM deep learning model with attention mechanism can increase the accuracy of short-term vegetable price predictions. Additionally, determine whether using China's economic policy uncertainty index and the Baidu index with extreme weather as the keyword can increase the accuracy of short-term vegetable price predictions. Firstly, the study uses Spearman correlation to rank the correlations of the variables. Secondly, it uses the TVP-VAR model to analyze the time-varying effects of the Baidu and China Economic Policy Uncertainty indices on the vegetable price index. Lastly, it uses the control variable method and fuses the CNN-LSTM-Attention model to test the effectiveness of the additional feature variables. According to the experimental results, RMSE, MAE, MAPE and R2 are 2.37, 1.78, 1.00% and 0.75, respectively. Benchmark research results indicate that the model's short-term prediction ability can be considerably enhanced by using the Baidu index and the China economic policy uncertainty index with extreme weather as the keywords.

Förderung

  • the Tianshan Talent Training Program
  • the Technology Innovation Project Fund of the Chinese Academy of Agriculture Sciences

Zitieren

APA

Bao, W., Su, W., Zhao, X., & Zhuang, J. (2025). A study on short-term vegetable price prediction based on the CNN-LSTM-Attention model. Discover Food, 5(1). https://doi.org/10.1007/s44187-025-00426-2

Vancouver

Bao W, Su W, Zhao X, Zhuang J. A study on short-term vegetable price prediction based on the CNN-LSTM-Attention model. Discover Food. 2025;5(1). doi:10.1007/s44187-025-00426-2

BibTeX
@article{Bao2025,
  author = {Bao, Weijia and Su, Wuzheng and Zhao, Xin and Zhuang, Jiayu},
  title = {A study on short-term vegetable price prediction based on the CNN-LSTM-Attention model},
  journal = {Discover Food},
  year = {2025},
  volume = {5},
  number = {1},
  doi = {10.1007/s44187-025-00426-2}
}
Kostenlos starten →