GNN-MAM: A graph neural network based multiple attention mechanism for regional financial risk prediction

By combining graph neural networks and multiple attention mechanisms, a GNN-MAM (Graph neural network based on multiple attention mechanisms) model was developed, which utilizes the structural characteristics of graph neural networks to capture complex correlations and dynamic changes in financial d...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuli Ma, MyeongCheol Choi, Yelin Weng
Format: Article
Language:English
Published: Elsevier 2025-08-01
Series:Alexandria Engineering Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1110016825007641
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:By combining graph neural networks and multiple attention mechanisms, a GNN-MAM (Graph neural network based on multiple attention mechanisms) model was developed, which utilizes the structural characteristics of graph neural networks to capture complex correlations and dynamic changes in financial data. Meanwhile, by introducing multiple attention mechanisms, the model can adaptively focus on key information and features in the data, thereby improving the accuracy and robustness of predictions. The experimental results show that compared with traditional financial risk prediction methods, GNN-MAM exhibits higher accuracy and robustness in regional financial risk prediction. Especially when dealing with datasets containing outliers, the predictive performance of GNN-MAM is significantly better than other methods, and the false positive rate is significantly reduced.
ISSN:1110-0168