A New Timescale For Fraud Science: Insights From Our CSO

By FeedZai


From the report, “At Feedzai, we recently completed a project: building a machine learning system to perform transaction fraud detection for a global payment service provider. This client does over a million transactions a day. We designed, trained, tested, and compared hundreds of models in just over a month. In the end we achieved a 16 percentage point increase in money recall with 20x fewer alerts. The fact that we’re continuing to deliver on projects of this scale, in this speed, makes me excited for the future. In this paper, I want to discuss why speed of data science matters, why we’re able to achieve it, and why it helps us deliver value for enterprises fighting fraud. ” Read on to find out more.

Topic Map