MAGNN: Multi-scale adaptive graph neural networks with contrastive learning for malicious network traffic detection

Puntos clave

• The approach demonstrates a notable increase in detection accuracy for malicious traffic.
• A key metric indicates a significant improvement in accuracy rates, surpassing previous models.
• Assessment using multi-scale adaptive graph neural networks enhanced by contrastive learning techniques.
• This method supports effective monitoring, highlighting the necessity of advanced algorithms in cybersecurity.