Towards more practical adversarial attacks on graph neural networks

Keyphrases

• Adversarial Attack 100%
• Attack Strategy 20%
• Attacker 20%
• Backward Propagation 20%
• Black Hole Attack 20%
• Black-box Settings 20%
• Classification Accuracy 60%
• Classification Loss 20%
• Common Graphs 20%
• Diminishing Returns 40%
• Effective Source 20%
• Gradient-based 20%
• Graph Neural Network 100%
• Inductive Bias 20%
• Large Margin 20%
• Loss Classification 20%
• Neural Network Model 20%
• Node Selection 20%
• PageRank 20%
• Random Walk 20%
• Real-time Constraints 20%
• Return Patterns 40%
• White-box Attack 20%

Computer Science

• Adversarial Machine Learning 100%
• Attackers 20%
• Black-Box Attack 20%
• Experimental Result 20%
• Graph Neural Network 100%
• Misclassification Rate 60%
• Neural Network Model 20%
• Random Walk 20%