一种基于自监督学习的图神经网络结构学习方法及系统
By employing a self-supervised learning method and utilizing multi-view generation and self-constraint mechanisms to optimize graph neural networks, the performance degradation caused by incomplete graph structure and noise, as well as the problem of smoothing node features, are solved, thus achieving more efficient graph neural network learning.
CN120911537BActive Publication Date: 2026-07-17XIAN AERONAUTICAL UNIV
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIAN AERONAUTICAL UNIV
- Filing Date
- 2025-07-22
- Publication Date
- 2026-07-17
Smart Images
Figure CN120911537B_ABST
Abstract
本发明涉及图神经网络技术领域，具体涉及一种基于自监督学习的图神经网络结构学习方法及系统，包括：获取节点特征矩阵和邻接矩阵；通过邻接转换器对邻接矩阵进行对称归一化处理，并进行多阶信息融合，获得融合邻接矩阵；通过节点特征矩阵获得二进制掩码矩阵，通过二进制掩码矩阵对节点特征矩阵进行扰动，获得新特征矩阵，通过新特征矩阵和融合邻接矩阵，获得新节点特征矩阵；通过融合邻接矩阵和新节点特征矩阵，确定出分类器和总损失函数；通过分类器和总损失函数进行自监督学习的图神经网络结构的学习。本发明提高了图神经网络结构学习的准确性。
Need to check novelty before this filing date? Find Prior Art