AI Paper Atlas

Research map / Graph learning / GNN architectures

Learning on Random Balls is Sufficient for Estimating (Some) Graph Parameters

Takanori Maehara, Hoang NT

NeurIPS 2021 · Graph learning › GNN architectures · approach: GNN generalization theory · 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
observe only random balls (55.7% of data) for train and test instead of full graphsGIN-style GNN · COLLABtest accuracy (%): 80.2 -> 80.3No clear effect
observe only random balls (14.9% memory) instead of full graphGIN-style GNN · RDT-BINARYtest accuracy (%): 92.4 -> 79.0Hurt
observe only random balls (14.2% memory) instead of full graphGIN-style GNN · RDT-MULTI5Ktest accuracy (%): 57.5 -> 44.0Hurt
RBS-GNN with partial observation vs WL subtree kernelRBS-GNN[≡2] · COLLABtest accuracy (%): 78.9 -> 80.3Improved
RBS-GNN with partial observation vs WL subtree kernelRBS-GNN[≡2] · RDT-BINARYtest accuracy (%): 81.0 -> 79.0Hurt
RBS-GNN with partial observation vs PatchySanRBS-GNN[≡2] · COLLABtest accuracy (%): 72.6 -> 80.3Improved
RBS-GNN with partial observation vs AWLRBS-GNN[≡2] · COLLABtest accuracy (%): 73.9 -> 80.3Improved
RBS-GNN with partial observation vs AWLRBS-GNN[≡2] · RDT-MULTI5Ktest accuracy (%): 54.7 -> 44.0Hurt
size generalization target: non-estimable max-degree instead of estimable global clustering coefficientRBS-GNN[≡2] · COLLABtest accuracy (%): 23.7 -> 16.1Hurt
size generalization target: non-estimable max-degree instead of estimable global clustering coefficientRBS-GNN[≡2] · RDT-BINARYtest accuracy (%): 33.7 -> 43.4Improved
size generalization target: non-estimable max-degree instead of estimable global clustering coefficientRBS-GNN[≡2] · RDT-MULTI5Ktest accuracy (%): 40.0 -> 39.8No clear effect

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