GNN Model Improves Cyberattack Detection with Minimal Data
/ 1 min read
🧩 New GNN-based model enhances cyberattack detection with minimal data. The paper introduces Few Edges Are Enough (FEAE), a Graph Neural Network (GNN) architecture that utilizes Self-Supervised Learning (SSL) and Few-Shot Learning (FSL) to improve cyberattack detection while minimizing reliance on labeled data. By employing a hybrid self-supervised objective that merges contrastive and reconstruction-based SSL techniques, FEAE effectively differentiates between false positives and actual attacks. Remarkably, the model demonstrates competitive performance using only one labeled malicious event per attack type, outperforming both self-supervised GNN baselines and some supervised methods on established network datasets. This advancement could significantly enhance the efficacy of attack detection systems in real-world scenarios where labeled data is scarce.
