Using machine learning to detect fraud, featuring Stephen Whitworth

Ravelin requires significant amounts of data from clients in order to provide the predictions as to which transactions are likely to result in chargebacks. In return, it is our commitment to our clients to use that data to the utmost degree in our machine learning for fraud detection, ensuring that we extract the most information possible to ensure we get the most predictions right.

It’s not perfect; perfection is an illusion in fraud detection. It is, however, optimal based on the premise that any online business wants to accept as many good transactions as possible.

In this podcast, Stephen Whitworth, Senior Data Engineer at Ravelin explains the basics of machine learning and discusses features, scores, and models.