A detector is a container that contains fraud detection logic, such as the models and rules, for one specific business event that you want to evaluate for fraud. You first create a detector by specifying the event that you have already defined and optionally add a model version that is already created and trained by Amazon Fraud Detector for the event.
You then add rules and rule execution order to a detector to create a version of the detector. A detector version defines the rules and optionally a model that will be run as part of the request for generating fraud predictions. You can add any of the rules defined within a detector to the detector version. You can also add any model trained on the evaluated event type to the detector version. A detector can have multiple versions, with each version having different rules and rule execution order to meet multiple use cases.
Each detector version must have a status of DRAFT, ACTIVE, or INACTIVE. Only one detector version can be in ACTIVE status at a time. Amazon Fraud Detector uses the detector version with ACTIVE status to generate fraud predictions.
