Improving Generalizability of Graph Anomaly Detection Models via Data Augmentation

Keyphrases

• Data Augmentation 100%
• Graph Anomaly Detection 100%
• Anomaly Detection Model 100%
• Training Data 33%
• Unseen 33%
• Anomaly Detection System 33%
• Training Model 16%
• Challenging Tasks 16%
• Superior Performance 16%
• Data Distribution 16%
• Unsupervised Method 16%
• Training Strategy 16%
• Detection Model 16%
• Testing Data 16%
• People's Needs 16%
• Improving Model 16%
• Poor Generalization 16%
• Networked Systems 16%
• Generalized Graph 16%
• Augmented Data 16%
• Test Graph 16%
• Limited Labels 16%
• Model Generalization 16%
• Domain Graph 16%
• Episodic Training 16%

Computer Science

• Data Augmentation 100%
• Anomaly Detection 100%
• Generalizability 100%
• Normal Distribution 33%
• Detection Method 33%
• Training Data 16%
• Data Distribution 16%
• Unsupervised Method 16%
• Superior Performance 16%