AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

MDGMIX: Boundary-Aware Subgraph Mixing for Multi-Domain Graph Pre-Training

Ziyu Zheng, Yaming Yang, Ziyu Guan, Wei Zhao, Xinyan Huang

ICML 2026 · Graph learning › Graph self-supervised learning · approach: Graph domain adaptation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add boundary-aware subgraph mixing and hierarchical discrimination to multi-domain graph pre-trainingMDGMIX · 1-shot node classification, 8 target graphsaccuracy: 43.38 -> 46.83 (Cora vs MDGPT)Improved
use boundary node selection instead of random subgraph mixup (ablation)MDGMIX · Cora node classificationaccuracy: 45.01 -> 46.83Improved
add coarse-grained domain classification loss (ablation)MDGMIX · Cora node classificationaccuracy: 44.22 -> 46.83Improved
add fine-grained domain decomposition loss (ablation)MDGMIX · Cora node classificationaccuracy: 43.15 -> 46.83Improved
add boundary mixup and domain decomposition together over domain classification alone (ablation)MDGMIX · Photo node classificationaccuracy: 62.79 -> 68.40Improved
keep domain decomposition when removing boundary mixup and domain classification (ablation)MDGMIX Variant 4 · Cora node classificationaccuracy: 45.28 vs 46.83 fullImproved
mix more than two domains via Dirichlet weights (ablation)MDGMIX · 7 target graphsaccuracy: Cora 46.20, 45.41, 45.06, 45.47 for 3-6 domainsNo clear effect
sample lambda from Beta distribution instead of fixed 0.5 (ablation)MDGMIX · Cora, Computers, Chameleonaccuracy: Computers 53.81 (Beta(0.2,0.2)) -> 54.95 (fixed 0.5)No clear effect
randomly drop up to 90% of nodes from pre-training graphs (ablation)multi-domain graph pre-training baselines · Pubmed, ComputersMicro accuracyNo clear effect
pre-train on boundary-centered mixed subgraphs and adapt with prompt weighting instead of full source graphs and domain tokensMDGMIX · pre-training and fine-tuning efficiencytime, memory, trainable parameters: fine-tuning parameters 6 vs 59 to 10240 for baselinesImproved
use MDGMIX on boundary subgraphs versus center or random subgraphs for domain classification (ablation)domain classifier · boundary analysisdomain classification accuracyImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy