Amazon Fraud Detector Documentation

Amazon Fraud Detector can automate the creation of machine learning models that identify potential fraud for common online activities such as new account creations, online payments, and guest checkouts. The automated model-building process is designed to perform various tasks such as data validation and enrichment, feature engineering, algorithm selection, hyperparameter tuning, and model deployment.  

Your model is designed to maintain its performance longer between trainings because Amazon Fraud Detector is designed to automatically calculate information like account age, time since last activity, and counts of activities. This means that your model can learn the difference between trusted customers who frequently make transactions and fraudsters’ continued attempts.

For models you train, you can see inputs you provided ranked by their impact on model performance. Using the importance values and relative ranking, you can gain insight into what inputs are driving your model performance.

Once you create an Amazon Fraud Detector fraud detection model, you can use the Amazon Fraud Detector console or application programming interface (API) to create rules based on model predictions. Customers can create rules to take actions such as accept, review, or collect more information for specific model scores. For example, you can easily create a rule to flag suspicious customer accounts for review if the model score is greater than your predetermined threshold and the account’s phone number country and IP address country do not match.

The Amazon Fraud Detector API is designed to perform real-time fraud predictions and evaluate online activities in your application as they occur. For example, you can call the fraud predictions API to check every new account sign-up for potential fraud risk, using your model and rules to trigger an action.

The Amazon Fraud Detector console is designed to allow easy search and review of your past fraud evaluations to audit detection logic. View event data, detection logic applied during the evaluation, and the conditions that resulted in a fraud prediction outcome.

Amazon Fraud Detection is designed to be integrated with Amazon SageMaker. You can use both your Amazon SageMaker and Amazon Fraud Detector models in your application to detect different types of fraud. For example, your application can use the Amazon Fraud Detector model to assess the fraud risk of customer accounts, and simultaneously use your Amazon SageMaker model to check for account compromise risk.

For additional information about service controls, security features and functionalities, including, as applicable, information about storing, retrieving, modifying, restricting, and deleting data, please see https://docs.aws.amazon.com/index.html. This additional information does not form part of the Documentation for purposes of the AWS Customer Agreement available at http://aws.amazon.com/agreement, or other agreement between you and AWS governing your use of AWS’s services.