AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Neural Link Prediction over Aligned Networks

Xuezhi Cao, Haokun Chen, Xuejian Wang, Weinan Zhang, Yong Yu

AAAI 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 19 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 all objective channels to direct links (ablation)MNN · FacebookAUC: 94.82% -> 96.96%Improved
add all objective channels to direct links (ablation)MNN · DoubanAUC: 95.84% -> 97.14%Improved
add multi-step link channels to direct links (ablation)MNN · FacebookAUC: 94.82% -> 95.95%Improved
add graphlet channel to direct links (ablation)MNN · TwitterAUC: 93.43% -> 93.71%No clear effect
add graphlet channel to direct links (ablation)MNN · FacebookAUC: 94.82% -> 95.20%Improved
stochastic instead of joint training (fixed batch count) (ablation)MNN · FacebookAUC: 95.11% -> 96.24%Improved
fixed batch count instead of fixed batch size (stochastic) (ablation)MNN · WeiboAUC: 95.19% -> 96.46%Improved
MNN with aligned-network channels vs best baselinelink prediction · Facebook 80% trainAUC: 93.39% -> 96.96%Improved
MNN vs best baseline under 20% training datalink prediction · WeiboAUC: 86.37% -> 92.39%Improved
MNN vs best baseline for degree-2 vertices (cold start)link prediction · cold startAUC: +8.53% relativeImproved
set beta weighting in [0.3, 0.5] (ablation)MNN · hyperparameterAUCImproved

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