Getting smart with cyber crime: Fighting fraud with machine learning – Featurespace

July 21, 2016

Getting smart with cyber crime: Fighting fraud with machine learning

How are companies fighting fraud with machine learning? Computer Business Review interviewed Martina King, CEO to find out.

With the UK government pledging to spend £1.9bn on cyber defences over the next five years CBR - Getting smart with cyber crime: Fighting fraud with machine learningand the UK’s National Crime Agency warning that cyber criminals remain ahead in technical capabilities, the advancements in machine learning technologies to fight fraud have never been more important.

Eleanor Burns talked to Martina about how fraud is being tackled using machine learning and adaptive behavioural analytics. The traditional approach has been to use fraud prevention systems based on the use of rules. These rules are effective for known attacks – it is possible to block the same type of attack again and again. However, this retrospective approach doesn’t help identify new attacks and leaves companies consistently one step behind.

Featurespace’s ARIC engine is different. It uses a self-learning mechanism, so the models never degrade and our customers in financial services, insurance and gaming can automatically adapt to new fraud types – this means that as criminals evolve, so does the system. At the same time, it works to build a profile of “normal” behaviour at an individual level, in real-time. It can then flag anomalous events for investigation, within the context of the change, to a high level of accuracy.

Read the full interview.