AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Graph Agreement Models for Semi-Supervised Learning

Otilia Stretcu, Krishnamurthy Viswanathan, Dana Movshovitz-Attias, Emmanouil Platanios, Sujith Ravi, Andrew Tomkins

NeurIPS 2019 · Graph learning › Graph self-supervised learning · approach: Semi-supervised node representation learning · 42 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 GAM regularizationMLP128 · Coraaccuracy (%): 51.7 -> 80.7Improved
add GAM regularizationGCN128 · Coraaccuracy (%): 80.9 -> 86.2Improved
add GAM regularizationGCN128 · Pubmedaccuracy (%): 76.9 -> 86.0Improved
replace NGM fixed edge weights with learned agreement (GAM)MLP128 · Coraaccuracy (%): 77.7 -> 80.7Improved
replace NGM with GAMGCN1024 · Pubmedaccuracy (%): 68.9 -> 81.6Improved
add GAM*, no graphMLP128 · Coraaccuracy (%): 51.7 -> 70.7Improved
add VATENTMLP128 · Coraaccuracy (%): 51.7 -> 24.1Hurt
add GAM* on CNNCNN · SVHNaccuracy (%): 72.33 -> 83.43Improved
add GAM* on top of VATCNN + VAT · CIFAR-10accuracy (%): 64.37 -> 69.64Improved
add GAM* on top of VATENTCNN + VATENT · CIFAR-10accuracy (%): 66.73 -> 67.29Mixed
use perfect agreement model instead of learned one (ablation)MLP128 + GAM · Coraaccuracy (%): 80.7 -> 90.5Improved
add NGM under noisy graph (<=60% correct edges) (ablation)MLP128 · Citeseeraccuracy (%)Hurt

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