Fraud & Risk
A New Timescale for Fraud Science
Machine learning systems are being used by a growing number of eCommerce companies to help mitigate fraud and risk. These systems, however, can require significant investments in time and resources to deploy, and complexity tends to increase considerably with very large data sets. This paper from Feedzai's CSO explores efforts to make data science both simpler and more scalable, then looks at ways to make data exploration easier. The article concludes with thoughts on possible future directions for machine learning, such as models "teaching" humans.
Two Heads Are Better Than One: How AI and HUMINT Converge to Battle Today's Sophisticated Online Fraudster