As a global payment processor with more than 2 billion cards in use around the world, Mastercard engages in a constant fight against fraud. The company processes around 165 million transactions per hour, and every one of those transactions must be examined in real time for signs of fraud.1
To accomplish this mind-boggling task, Mastercard relies on the power of high performance computing (HPC) systems driving machine learning algorithms. These algorithms apply 1.9 million rules to each transaction in a matter of milliseconds. These rules examine things like the cardholders’ buying habits, geographic locations and travel patterns, along with real-time data on card usage — such as what they are trying to buy and where they are trying to buy it.1
None of this would be possible without machine learning algorithms.
So, what is machine learning?
In a few words, machine learning is simply the process of training a system by feeding large amounts of data into an algorithm to help the system learn how to perform a task. Machine learning is one of the most fundamental building blocks for the AI solutions used in the fight against fraud, cybercrime and similar threats to the enterprise.
Let’s look at some examples of how enterprises are using machine learning at ever-larger scales to ward off fraud and cybercrime.
- In the retail industry, merchants use machine learning and deep learning — a form of machine learning — to protect the security and privacy of shoppers’ personal information and credit cards. At the same time, merchants use machine learning to gain insights into customers’ buying habits and behaviors — insights that help merchants offer their customers the right products at the right time.
- In the financial services industry, companies use machine learning techniques to detect signs of fraud and cybercrime. Machine learning helps financial services companies flag suspicious behavior as it occurs and build threat-detection systems that learn from their experiences and improve over time, while gathering valuable insights into customers’ needs and interests.
- The insurance industry uses AI-driven solutions to monitor claims for fraud, using tools like Intel® Saffron™ AI to analyze datasets to find hidden patterns, trends and similarities. In one case, an auto insurance company used Intel Saffron AI to examine 113,000 claims from one year in one state in less than a month. By highlighting hidden connections among the clinics, this AI-driven examination uncovered three potential fraud rings involving a radiology clinic and other medical providers.2
The horsepower under the hood
It takes a lot of processing power, high performance storage and fast networking to enable machine learning algorithms to work their magic in real time. To meet this need, companies are deploying Dell EMC Ready Solutions for AI. These pre-designed and pre-validated solutions bring together just about everything businesses need to accelerate their AI initiatives — and go from AI-possible to AI-ready.
The growing portfolio of Ready Solutions includes the new Dell EMC Ready Solutions for AI – Deep Learning with Intel. It’s a scalable, flexible container-based solution for the development and testing of deep learning and machine learning models using Intel® Xeon® Scalable processors. Co-developed by Dell EMC and Intel, the solution brings together a high compute-density rackmount server chassis, Ethernet networking, expandable hybrid disk/flash storage, optimized libraries and deep learning frameworks to enable organizations to train AI models quickly and efficiently.
Machine and deep learning solutions are now on the frontlines in the battle against fraud and cybercrime in industries ranging from retail and payments processing to financial services and insurance. And with new Ready Solutions that make AI systems easier to deploy, IT teams now have easier access to the weapons they need to fight back in a manner that better protects the enterprise and its customers.
To learn more
Advancing the Frontiers of AI
Dramatic advances in data analytics and high performance computing capabilities have created a foundation for the adoption of AI-driven applications in the enterprise. But these enabling technologies are only part of the AI story. The other part is the rise of smarter algorithms that can glean insights from massive amounts of data. In this series, we explore these building blocks for AI solutions in enterprise environments.
1 Dell EMC, “Fighting fraud the smart way — with data analytics and artificial intelligence,” December 2018.
2 Intel, “AI’s Role in Fighting Fraud,” November 9, 2017.