Unsupervised Machine Learning: Fighting Cross-Channel Fraud
ThetaRay’s James Heinzman interviewed by Bankinfosecurity.com and shares insights on countering sophisticated fraud techniques
Attackers are exploiting cyber gaps and combining them with other fraud channels. Conventional rule-based defensive techniques and supervised machine learning-based methods are not very good at detecting these sophisticated attacks, says Heinzman. Attackers are beginning to study and mimic employee behavior in complex ways, and they have access on the dark web to security playbooks for older rules-based and supervised strategies – which puts the defender several steps behind, he contends.