Artificial intelligence and machine learning are becoming increasingly important tools in payment fraud detection as criminals develop more sophisticated methods of attacking accounts and transactions. However, building effective fraud models traditionally requires data scientists to manually identify the features and behavioral patterns most likely to indicate suspicious activity.
Featurespace is seeking to advance payment fraud detection through Automated Deep Behavioral Networks on its ARIC Risk Hub platform. By automating feature discovery and improving the system’s understanding of transaction timing and behavior, the technology is designed to identify changing fraud patterns more quickly and help financial institutions prevent fraudulent transactions before funds leave an account.
Featurespace claims to have implemented a breakthrough on its ARIC Risk Hub platform. By automating the identification of feature selection the platform can monitor and learn fraud patterns faster and apply the solution to a broader set of fraud related problems:
“Deep learning technology has various applications, such as in natural language processing for the prediction of the next word in a sentence, however its use in preventing fraud in card and payments fraud detection has not been optimised to protect companies and consumers from card and payments fraud. With this invention, that challenge is solved.
Transactions are intermittent, making contextual understanding of time critical to predicting behaviour. Previously, building effective machine-learning models for fraud prevention required data scientists to have deep domain expertise to identify and select appropriate data features – a laborious, yet vital step.
Featurespace research developed Automated Deep Behavioural Networks to automate feature discovery and introduce memory cells with native understanding of the significance of time in transaction flows, improving upon the market-leading performance of the company’s Adaptive Behavioural Analytics. Detecting fraud before the victim’s money leaves the account is the best line of defence against scams, account takeover, card and payment fraud attacks.”
Automating feature discovery could significantly improve how financial institutions approach payment fraud detection. Instead of relying as heavily on data scientists to manually identify the characteristics that indicate fraud, deep learning models can continuously evaluate transaction behavior and uncover relevant patterns.
The ability to understand the importance of time within transaction flows is particularly valuable because fraudulent activity can be difficult to identify from an individual transaction alone. Evaluating transactions within the context of a customer’s broader behavior can provide stronger signals of scams, account takeover, and card fraud.
As fraud tactics continue to evolve, technologies that learn and adapt more quickly can give financial institutions another layer of protection. Featurespace’s Automated Deep Behavioral Networks represent an effort to make sophisticated behavioral analytics more scalable while improving the speed and accuracy of fraud detection.
Overview by Tim Sloane, VP, Payments Innovation at Mercator Advisory Group
