AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Improving the Robustness of Wasserstein Embedding by Adversarial PAC-Bayesian Learning

Daizong Ding, Mi Zhang, Xudong Pan, Min Yang, Xiangnan He

AAAI 2020 · Adversarial robustness and security › Attacks on text and graphs · approach: Graph neural network robustness · 2 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
RAWEN instead of best baseline DVNEnode embedding · Google link predictionprecision: 5%Improved
adversarial PAC-Bayesian training (RAWEN) under 18% edge noisenode embedding · Wiki-Vote link prediction, noise level 6AUC: 2% decrease vs 10% decreaseImproved
distribution-based embedding (DVNE) instead of vector-based embeddingsnode embedding baselines · link predictionprecision and recallImproved
add adversarial learning (GANE) to vector-based embeddingvector-based embedding · link predictionprecision and recallImproved
use Gaussian instead of Dirichlet or truncated Gaussian embedding distribution (ablation)RAWEN · link predictionprecision and recallImproved
choice of embedding distribution (Gaussian, Dirichlet, truncated Gaussian) (ablation)RAWEN · node classification (Email, Wiki)Macro/Micro-F1No clear effect
use Dirichlet distribution for embeddings (ablation)RAWEN · t-SNE visualizationclustering structure (qualitative)Improved

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