AI Paper Atlas

Research map / Graph learning / GNN architectures

Wasserstein Graph Neural Networks for Graphs With Missing Attributes

Zhixian Chen, Tengfei Ma, Yangqiu Song, Yang Wang

TPAMI 2025 · Graph learning › GNN architectures · approach: Graph structure learning · 6 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 low-dimensional SVD distributional embedding instead of high-dimensional attribute distributions (ablation)WGNN_prob · synthetic graph, high missing ratesaccuracyImproved
use SVD latent factors instead of random orthogonal latent factors (ablation)WGNN_rand · synthetic graph, all missing settingsaccuracyImproved
add low-dimensional embedding instead of direct high-dimensional distributions (ablation)WGNN_hd · Cora, r_p=0.1accuracy: 0.439 -> below 0.6029 GCN_nofeatImproved
use Wasserstein aggregation instead of GCN aggregation on U_k (ablation)GCN_U · Cora, partially missingaccuracyImproved
use Wasserstein aggregation instead of Mean aggregation on U_k (ablation)MEAN_U · Cora, partially missingaccuracyMixed
add Euclidean embedding with V_k (Lambda_k V_k^T) before prediction (ablation)WGNN_svd,U · Cora, partially missingaccuracyImproved
add Wasserstein aggregation with SVD embedding at highest entirely missing levelGCN and other imputation baselines · Cora/Citeseer/PubMed, r_e=0.9accuracy: at least 6%, 10% and 9% higher accuracyImproved
add Wasserstein aggregation with SVD embedding at highest partially missing levelGCN and other imputation baselines · Cora/Citeseer/PubMed, r_p=0.9accuracy: 5%, 8% and 5% higher accuracyImproved
give MIDA 30% more observed valuesMIDA baseline · Cora/Citeseer/PubMedaccuracyNo clear effect
increase layers L and aggregation steps h (up to L=5, h=9)WGNN_svd · Cora, r_e=0.9accuracyImproved
use Multi-WGNN instead of IGMC for matrix completionIGMC · FlixsterRMSE: 0.872 -> 0.883No clear effect
use Multi-WGNN instead of IGMC for matrix completionIGMC · MovieLens-100KRMSE: 0.905 -> 0.910No 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