We’re always working to find new ways to protect customers and prevent successful attacks, and one recent addition to our research arsenal is the use of unsupervised machine learning on large datasets of domain information. Machine learning-based techniques like this can help us discover new threats and block them before they can affect our customers. They can quickly identify malicious domains that are part of larger campaigns as soon as they become active and provide much broader coverage for these campaigns than traditional methods.

https://researchcenter.paloaltonetworks.com/2018/10/unit42-detecting-malicious-campaigns-machine-learning/