Aug 2026
Dynamic pseudo-label guided adversarial multi-scale graph convolutional network for cross-domain fault diagnosis.
A Dynamic Pseudo-Label Guided Adversarial Multi-Scale Graph Convolutional Network (DPAMGCN) for unsupervised cross-domain fault diagnosis and a dynamic threshold-based pseudo-label filtering strategy is proposed that enhances the model's generalization capability.
Jinqi Gao, Bo Zhang, Tianlong Huo et al.
· Review of Scientific Instrum... · 0 citations