Correntropy-Induced Wasserstein GCN: Learning Graph Embedding via Domain Adaptation

Abstract

We propose a novel Graph Convolutional Network (GCN) that leverages correntropy-induced loss functions and Wasserstein distance for robust domain adaptation in graph embedding tasks.

BibTeX

@article{wang2023cwrgcn,
  title={Correntropy-Induced Wasserstein GCN: Learning Graph Embedding via Domain Adaptation},
  author={Wang, Wei and Zhang, Gaowei and Han, Hongyong and Zhang, Chi},
  journal={IEEE Transactions on Image Processing},
  volume={32},
  pages={3980--3993},
  year={2023},
  publisher={IEEE}
}