Abstract:Existing machine learning literature lacks graph-based domain adaptation techniques capable of handling large distribution shifts, primarily due to the difficulty in simulating a coherent evolutionary path from source to target graph. To meet this challenge, we present a graph gradual domain adaptation (GGDA) framework, which constructs a compact domain sequence that minimizes information loss during adaptation. Our approach starts with an efficient generation of knowledge-preserving intermediate graphs over the Fused Gromov-Wasserstein (FGW) metric. A GGDA domain sequence is then constructed upon this bridging data pool through a novel vertex-based progression, which involves selecting "close" vertices and performing adaptive domain advancement to enhance inter-domain transferability. Theoretically, our framework provides implementable upper and lower bounds for the intractable inter-domain Wasserstein distance, $W_p(\mu_t,\mu_{t+1})$, enabling its flexible adjustment for optimal domain formation. Extensive experiments across diverse transfer scenarios demonstrate the superior performance of our GGDA framework.
| Comments: | Accepted by ACM Trans. Intell. Syst. Technol. (this https URL) |
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2501.17443 [cs.LG] |
| (or arXiv:2501.17443v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2501.17443 arXiv-issued DOI via DataCite |
Submission history
From: Pui Ieng Lei [view email]
[v1]
Wed, 29 Jan 2025 06:48:59 UTC (3,382 KB)
[v2]
Fri, 27 Jun 2025 14:45:02 UTC (1,566 KB)
[v3]
Thu, 28 Aug 2025 11:01:38 UTC (1,762 KB)
[v4]
Wed, 13 May 2026 04:42:02 UTC (3,281 KB)