Amazon launches machine learning-based fraud detector

Amazon Web Services (AWS) has announced the general availability of its machine learning-based fraud detection service.

advertisement

Detect fraud faster

Amazon Fraud Detector is a fully managed service that aims to facilitate the rapid identification of potentially fraudulent online activity, such as fraudulent online payment and identity theft, creation of fake accounts and abuse of loyalty accounts and promotion codes "in milliseconds.

To use the service, users Customers can select a pre-built machine learning template, upload data on historical transaction events - both fraudulent and legitimate - to build, train, and deploy machine learning models, and create decision logic to attribute results to forecasts.

"Customers of all sizes and from all industries have trusted ussaid they spend a lot of time and effort trying to reduce the number of frauds on their websites and applications, says Swami Sivasubramanian, vice president of Amazon Machine Learning. "Drawing on 20 years of fraud detection experience and powerful machine learning technology, we are delighted to offer our customers the Amazon Fraud Detector, which enables them to automatically detect potential fraud. , save time and money, and improve the customer experience, without the need for a machine learning experience.

Building a Fraud Detection Model

Depending on the type of fraud customers want to predict, Amazon Fraud Detector will pre-process the data, select a algorithm and will train a model.

Amazon Fraud Detector trains and deploys a model to a fully managed private API endpoint.

Customers can send againvile API activities, such as signings or new purchases, to receive a fraud risk response, which includes a fraud risk score. Based on this report, a customer's application can determine the correct course of action, for example, to accept a purchase or pass it to a human for review, AWS details.

Internal services then deployed for Amazon customers

This service was announced at the AWS re: Invent conference in December. At the time, CEO Andy Jassy explained that the company was always looking at what it can offer, including within its internal capabilities, to deliver value to its customers.

"A number of services have been deployed internally at Amazon, which customers have asked us to offer them as a service," he adds, citing Amazon Lex, Amazon Personalize, and Amazon Forecast as recent examples.

"We do detection of fsteep for over 20 years.

Machine learning is very effective, but difficult to implement

Andy Jassy explains that machine learning is "incredibly useful but it is difficult for most companies to use it to detect fraud.

"You send us transaction data - email addresses, IP addresses, maybe phone numbers , as well as fraudulent and legitimate transactions - then we aggregate that data with the algorithms we've built ... and we build a unique model for you, he explains. "Then it's exposed through an API.

"This is a completely different way of dealing with fraud through machine learning," he adds.

Amazon Fraud Detector is available today in parts of the United States, Ireland, Singapore and Sydney. It should be available ins other regions in the coming months.

Source: