Graph Harmony: Denoising and Nuclear-Norm Wasserstein Adaptation for Enhanced Domain Transfer in Graph-Structured Data
Abstract: Graph-structured data can be found in numerous domains, yet the scarcity of labeled instances hinders its effective utilization of deep learning in many scenarios. Traditional unsupervised domain adaptation (UDA) strategies for graphs primarily hinge on adversarial learning and pseudo-labeling. These approaches fail to effectively leverage graph discriminative features, leading to class mismatching and unreliable label quality. To address these obstacles, we developed the Denoising and Nuclear-Norm Wasserstein Adaptation Network (DNAN). DNAN employs the Nuclear-norm Wasserstein discrepancy (NWD), which can simultaneously achieve domain alignment and class distinction. It also integrates a denoising mechanism via a Variational Graph Autoencoder. This denoising mechanism helps capture essential features of both source and target domains, improving the robustness of the domain adaptation process. Our comprehensive experiments demonstrate that DNAN outperforms state-of-the-art methods on standard UDA benchmarks for graph classification.
Submission Length: Long submission (more than 12 pages of main content)
Previous TMLR Submission Url: https://openreview.net/forum?id=8HmMzL2rm3&referrer=%5Bthe%20profile%20of%20Mengxi%20Wu%5D(%2Fprofile%3Fid%3D~Mengxi_Wu1)
Changes Since Last Submission: We are resubmitting our paper based on the final comment of the AE who supervised the review of our initial submission and encouraged us to resubmit our work. Compared to the previous version, we changed the format of our paper from a short paper to a long paper due to the substantial additions we had. We also addressed the concerns raised by previous reviewers. Please refer to the cover letter in the supplementary materials zipped folder about the details of our changes. In the main paper, we marked the portion we added in blue.
Assigned Action Editor: ~Weijian_Deng1
Submission Number: 2535
Loading