vpnMentor Interview: Using Multi-Dimensional Algos to Detect Fraud

June 22, 2018

DataVisor CEO - Yinglian Xie

DataVisor’s CEO Yinglian Xie was recently interviewed by Gail Rand of vpnMentor about DataVisor’s proprietary, multi-dimensional Unsupervised Machine Learning algorithms that are used protect over 4 billion accounts from fraud and abuse at some of the largest internet and financial properties in the world.

The interview covers Yingilan’s background at Carnegie Mellon and Microsoft, how she met her co-founder Fang Yu and how they decided to found DataVisor together. It continues to explore biggest fraud and abuse threats to social commerce and financial companies and how attackers use sophisticated techniques to avoid incubation.

Yinglian also discusses DataVisor’s proprietary Unsupervised Machine Learning Engine, which analyzes the subtle connections between accounts, rather than looking at each account in isolation. This patented system allows DataVisor to detect armies of malicious accounts created and orchestrated by attackers to launch their attacks. Because these hidden connections between accounts exist prior to the actual attack, the UML Engine can often detect attackers before they have even launched their attacks.

Learn more about DataVisor and Unsupervised Machine Learning at vpnMentor >>