Argyle Data, a leader in big data/machine learning analytics for mobile providers, has highlighted the role of supervised and unsupervised machine learning in detecting and preventing anomalous mobile traffic. The move comes as Argyle Data and Carnegie Mellon University (CMU) Silicon Valley’s Department of Electrical and Computer Engineering prepare to publish a new research paper on anomaly detection, which will be presented at academic conferences during the first half of 2017.
Global mobile fraud levels cost the industry an estimated U.S. $38 billion 2015 according to the latest CFCA survey. Most major attacks today are ‘fraud cocktails’: unpredictable mixtures of several fraud types.
The chief reason that operators are unable to detect complex new fraud is that approaches currently used to detect fraud in communications networks typically rely on static rules with pre-set thresholds, and can only detect known fraud types. Modern cyber-attacks evolve faster than an analyst can write rules to detect them. This type of mutating attack happens on a massive scale, siphoning off millions of dollars in just a few minutes, and is impervious to detection with traditional methods.
Rapid identification of suspicious traffic is the ultimate goal of telco fraud analysts but again, most existing fraud prevention technologies require time-consuming manual review resulting in costly delays in remediation.
There is a growing need for new fraud detection solutions that can adapt to evolving network crime and usage patterns and all the signs point to machine learning as the answer,” said Mohan Gyani, member of the Argyle Data Board of Directors and former president and CEO of AT&T Wireless Mobility. “Facebook, Google, and LinkedIn have pioneered big data and machine learning approaches to protecting their subscribers and gaining insight on vast amounts of data. The way in which Argyle Data is applying its supervised and unsupervised machine learning techniques to solve fraud in the mobile industry is highly innovative. Their upcoming research with CMU will provide further evidence of their leadership in this field.”
Argyle Data is used by the world’s leading mobile operators to detect fraud, profit, and SLA threats that cost the industry $38 billion per year. Argyle Data’s industry-leading native Hadoop application suite uses the latest machine learning technologies against a unique, comprehensive data lake to give communications service providers a 360-degree view of user activities, allowing them to detect in real time the previously undiscoverable revenue threats and attack patterns being waged against their networks.
Not all anomalies are fraud, but all fraud is anomalous. The only way to detect anomalies in real time is to apply machine learning at massive scale,” commented Robbie Lynch of telecommunications consulting firm Robbie Lynch & Associates, a member of the GSMA Fraud & Security Group, and former Fraud & Security Manager at Vodafone. “This was not possible until the advent of unsupervised machine learning, which makes it possible to analyze massive amounts of data in seconds and present instant alerts about anomalous traffic to fraud analysts. Argyle Data’s big data machine learning system is meeting a clear need for innovation in this area.”
Sign up for the free insideAI News newsletter.
