Build an architecture that uses Amazon SageMaker to detect potentially fraudulent activity and flag that activity for review

Publication date: May 2019 (last update: January 2022)

Fraud is an ongoing problem that can cost businesses billions of dollars annually and damage customer trust. Many companies use a rule-based approach to detect fraudulent activity where fraud patterns are defined as rules. But implementing and maintaining rules can be a complex, time-consuming process because fraud is constantly evolving, rules require fraud patterns to be known, and rules can lead to false positives or false negatives.

Machine learning (ML) provides a flexible approach to fraud detection. ML models do not use pre-defined rules to determine whether activity is fraudulent. Instead, they are trained to recognize fraud patterns in datasets, and the models are self-learning, which allows them to adapt to new, unknown fraud patterns. In addition, unsupervised ML models allow us to extract knowledge from unlabeled data, flagging anomalous transactions for review.

This solution automates the detection of potentially fraudulent activity, and flags that activity for review. To help you build, train, and deploy ML models at scale, Fraud Detection Using Machine Learning leverages Amazon SageMaker for real-time fraud detection. It also includes an example dataset, but you can modify the solution to work with any dataset.

This implementation guide discusses architectural considerations and configuration steps for deploying Fraud Detection Using Machine Learning on the Amazon Web Services (AWS) Cloud. It includes links to an AWS CloudFormation template that launches and configures the AWS services required to deploy this solution using AWS best practices for security and availability.

The guide is intended for developers and data scientists who have practical experience with machine learning and architecting on the AWS Cloud.