Attacking Graph-Based Classification without Changing Existing Connections

Keyphrases

• Attacker 100%
• Fake Nodes 100%
• Graph Classification 100%
• Classification Methods 66%
• Node Set 66%
• Real Social Network 66%
• Graph Data 66%
• Machine Learning 33%
• Rapid Development 33%
• Real-world Application 33%
• Optimization Problem 33%
• Adversarial Attack 33%
• Adversary 33%
• Machine Learning Models 33%
• Adversarial Learning 33%
• Gradient Method 33%
• False Negative Rate 33%
• Adjacency Matrix 33%
• New Nodes 33%
• Collective Classification 33%
• Novel Attacks 33%
• Fake Accounts 33%
• Feature Matrix 33%
• Evade 33%
• Machine Learning Study 33%

Computer Science

• Attackers 100%
• Adversarial Machine Learning 66%
• Classification Method 66%
• Social Network 66%
• Optimization Problem 33%
• World Application 33%
• Rapid Development 33%
• False Negative 33%
• Machine Learning 33%
• Learning System 33%
• Adjacency Matrix 33%
• Feature Matrix 33%
• Machine Learning Model 33%