List<E> eventVariableNames
The names of all the event variables that were used to derive the aggregated variables.
String relativeImpact
The relative impact of the aggregated variables in terms of magnitude on the prediction scores.
Float logOddsImpact
The raw, uninterpreted value represented as log-odds of the fraud. These values are usually between -10 to +10, but range from -infinity to +infinity.
A positive value indicates that the variables drove the risk score up.
A negative value indicates that the variables drove the risk score down.
String name
The name of the list.
String description
The description of the list.
String variableType
The variable type of the list.
String createdTime
The time the list was created.
String updatedTime
The time the list was last updated.
String arn
The ARN of the list.
Float cr
The challenge rate. This indicates the percentage of login events that the model recommends to challenge such as one-time password, multi-factor authentication, and investigations.
Float adr
The anomaly discovery rate. This metric quantifies the percentage of anomalies that can be detected by the model at the selected score threshold. A lower score threshold increases the percentage of anomalies captured by the model, but would also require challenging a larger percentage of login events, leading to a higher customer friction.
Float threshold
The model's threshold that specifies an acceptable fraud capture rate. For example, a threshold of 500 means any model score 500 or above is labeled as fraud.
Float atodr
The account takeover discovery rate. This metric quantifies the percentage of account compromise events that can be detected by the model at the selected score threshold. This metric is only available if 50 or more entities with at-least one labeled account takeover event is present in the ingested dataset.
Float asi
The anomaly separation index (ASI) score. This metric summarizes the overall ability of the model to separate anomalous activities from the normal behavior. Depending on the business, a large fraction of these anomalous activities can be malicious and correspond to the account takeover attacks. A model with no separability power will have the lowest possible ASI score of 0.5, whereas the a model with a high separability power will have the highest possible ASI score of 1.0
List<E> metricDataPoints
The model's performance metrics data points.
ATIModelPerformance modelPerformance
The model's overall performance scores.
String jobId
The ID of the batch import job.
String status
The status of the batch import job.
String failureReason
The reason batch import job failed.
String startTime
Timestamp of when the batch import job started.
String completionTime
Timestamp of when batch import job completed.
String inputPath
The Amazon S3 location of your data file for batch import.
String outputPath
The Amazon S3 location of your output file.
String eventTypeName
The name of the event type.
String iamRoleArn
The ARN of the IAM role to use for this job request.
String arn
The ARN of the batch import job.
Integer processedRecordsCount
The number of records processed by batch import job.
Integer failedRecordsCount
The number of records that failed to import.
Integer totalRecordsCount
The total number of records in the batch import job.
String jobId
The job ID for the batch prediction.
String status
The batch prediction status.
String failureReason
The reason a batch prediction job failed.
String startTime
Timestamp of when the batch prediction job started.
String completionTime
Timestamp of when the batch prediction job completed.
String lastHeartbeatTime
Timestamp of most recent heartbeat indicating the batch prediction job was making progress.
String inputPath
The Amazon S3 location of your training file.
String outputPath
The Amazon S3 location of your output file.
String eventTypeName
The name of the event type.
String detectorName
The name of the detector.
String detectorVersion
The detector version.
String iamRoleArn
The ARN of the IAM role to use for this job request.
String arn
The ARN of batch prediction job.
Integer processedRecordsCount
The number of records processed by the batch prediction job.
Integer totalRecordsCount
The total number of records in the batch prediction job.
String jobId
The ID of an in-progress batch import job to cancel.
Amazon Fraud Detector will throw an error if the batch import job is in FAILED,
CANCELED, or COMPLETED state.
String jobId
The ID of the batch prediction job to cancel.
String jobId
The ID of the batch import job. The ID cannot be of a past job, unless the job exists in
CREATE_FAILED state.
String inputPath
The URI that points to the Amazon S3 location of your data file.
String outputPath
The URI that points to the Amazon S3 location for storing your results.
String eventTypeName
The name of the event type.
String iamRoleArn
The ARN of the IAM role created for Amazon S3 bucket that holds your data file.
The IAM role must have read permissions to your input S3 bucket and write permissions to your output S3 bucket. For more information about bucket permissions, see User policy examples in the Amazon S3 User Guide.
List<E> tags
A collection of key-value pairs associated with this request.
String jobId
The ID of the batch prediction job.
String inputPath
The Amazon S3 location of your training file.
String outputPath
The Amazon S3 location of your output file.
String eventTypeName
The name of the event type.
String detectorName
The name of the detector.
String detectorVersion
The detector version.
String iamRoleArn
The ARN of the IAM role to use for this job request.
The IAM Role must have read permissions to your input S3 bucket and write permissions to your output S3 bucket. For more information about bucket permissions, see User policy examples in the Amazon S3 User Guide.
List<E> tags
A collection of key and value pairs.
String detectorId
The ID of the detector under which you want to create a new version.
String description
The description of the detector version.
List<E> externalModelEndpoints
The Amazon Sagemaker model endpoints to include in the detector version.
List<E> rules
The rules to include in the detector version.
List<E> modelVersions
The model versions to include in the detector version.
String ruleExecutionMode
The rule execution mode for the rules included in the detector version.
You can define and edit the rule mode at the detector version level, when it is in draft status.
If you specify FIRST_MATCHED, Amazon Fraud Detector evaluates rules sequentially, first to last,
stopping at the first matched rule. Amazon Fraud dectector then provides the outcomes for that single rule.
If you specifiy ALL_MATCHED, Amazon Fraud Detector evaluates all rules and returns the outcomes for
all matched rules.
The default behavior is FIRST_MATCHED.
List<E> tags
A collection of key and value pairs.
String name
The name of the list.
List<E> elements
The names of the elements, if providing. You can also create an empty list and add elements later using the UpdateList API.
String variableType
The variable type of the list. You can only assign the variable type with String data type. For more information, see Variable types.
String description
The description of the list.
List<E> tags
A collection of the key and value pairs.
String modelId
The model ID.
String modelType
The model type.
String trainingDataSource
The training data source location in Amazon S3.
TrainingDataSchema trainingDataSchema
The training data schema.
ExternalEventsDetail externalEventsDetail
Details of the external events data used for model version training. Required if trainingDataSource
is EXTERNAL_EVENTS.
IngestedEventsDetail ingestedEventsDetail
Details of the ingested events data used for model version training. Required if trainingDataSource
is INGESTED_EVENTS.
List<E> tags
A collection of key and value pairs.
String ruleId
The rule ID.
String detectorId
The detector ID for the rule's parent detector.
String description
The rule description.
String expression
The rule expression.
String language
The language of the rule.
List<E> outcomes
The outcome or outcomes returned when the rule expression matches.
List<E> tags
A collection of key and value pairs.
Rule rule
The created rule.
String name
The name of the variable.
String dataType
The data type of the variable.
String dataSource
The source of the data.
String defaultValue
The default value for the variable when no value is received.
String description
The description.
String variableType
The variable type. For more information see Variable types.
Valid Values:
AUTH_CODE | AVS | BILLING_ADDRESS_L1 | BILLING_ADDRESS_L2 | BILLING_CITY | BILLING_COUNTRY | BILLING_NAME | BILLING_PHONE | BILLING_STATE | BILLING_ZIP | CARD_BIN | CATEGORICAL | CURRENCY_CODE | EMAIL_ADDRESS | FINGERPRINT | FRAUD_LABEL | FREE_FORM_TEXT | IP_ADDRESS | NUMERIC | ORDER_ID | PAYMENT_TYPE | PHONE_NUMBER | PRICE | PRODUCT_CATEGORY | SHIPPING_ADDRESS_L1 | SHIPPING_ADDRESS_L2 | SHIPPING_CITY | SHIPPING_COUNTRY | SHIPPING_NAME | SHIPPING_PHONE | SHIPPING_STATE | SHIPPING_ZIP | USERAGENT
List<E> tags
A collection of key and value pairs.
String jobId
The ID of the batch import job to delete.
String jobId
The ID of the batch prediction job to delete.
String detectorId
The ID of the detector to delete.
String name
The name of the entity type to delete.
String eventTypeName
The name of the event type.
String name
The name of the event type to delete.
String modelEndpoint
The endpoint of the Amazon Sagemaker model to delete.
String name
The name of the label to delete.
String name
The name of the list to delete.
String name
The name of the outcome to delete.
Rule rule
String name
The name of the variable to delete.
String detectorId
The detector ID.
String description
The detector description.
String eventTypeName
The name of the event type.
String lastUpdatedTime
Timestamp of when the detector was last updated.
String createdTime
Timestamp of when the detector was created.
String arn
The detector ARN.
String name
The entity type name.
String description
The entity type description.
String lastUpdatedTime
Timestamp of when the entity type was last updated.
String createdTime
Timestamp of when the entity type was created.
String arn
The entity type ARN.
String modelEndpoint
The endpoint of the external (Amazon Sagemaker) model.
Boolean useEventVariables
Indicates whether event variables were used to generate predictions.
Map<K,V> inputVariables
Input variables use for generating predictions.
Map<K,V> outputVariables
Output variables.
String ruleId
The rule ID.
String ruleVersion
The rule version.
String expression
The rule expression.
String expressionWithValues
The rule expression value.
List<E> outcomes
The rule outcome.
Boolean evaluated
Indicates whether the rule was evaluated.
Boolean matched
Indicates whether the rule matched.
String eventId
The event ID.
String eventTypeName
The event type.
String eventTimestamp
The timestamp that defines when the event under evaluation occurred. The timestamp must be specified using ISO 8601 standard in UTC.
Map<K,V> eventVariables
Names of the event type's variables you defined in Amazon Fraud Detector to represent data elements and their corresponding values for the event you are sending for evaluation.
String currentLabel
The label associated with the event.
String labelTimestamp
The timestamp associated with the label to update. The timestamp must be specified using ISO 8601 standard in UTC.
List<E> entities
The event entities.
Boolean eventBridgeEnabled
Specifies if event orchestration is enabled through Amazon EventBridge.
String eventId
The event ID.
String eventTypeName
The event type.
String eventTimestamp
The timestamp of the event.
String predictionTimestamp
The timestamp when the prediction was generated.
String detectorId
The detector ID.
String detectorVersionId
The detector version ID.
String name
The event type name.
String description
The event type description.
List<E> eventVariables
The event type event variables.
List<E> labels
The event type labels.
List<E> entityTypes
The event type entity types.
String eventIngestion
If Enabled, Amazon Fraud Detector stores event data when you generate a prediction and uses that
data to update calculated variables in near real-time. Amazon Fraud Detector uses this data, known as
INGESTED_EVENTS, to train your model and improve fraud predictions.
IngestedEventStatistics ingestedEventStatistics
Data about the stored events.
String lastUpdatedTime
Timestamp of when the event type was last updated.
String createdTime
Timestamp of when the event type was created.
String arn
The entity type ARN.
EventOrchestration eventOrchestration
The event orchestration status.
String modelEndpoint
The Amazon SageMaker model endpoints.
String modelSource
The source of the model.
String invokeModelEndpointRoleArn
The role used to invoke the model.
ModelInputConfiguration inputConfiguration
The input configuration.
ModelOutputConfiguration outputConfiguration
The output configuration.
String modelEndpointStatus
The Amazon Fraud Detector status for the external model endpoint
String lastUpdatedTime
Timestamp of when the model was last updated.
String createdTime
Timestamp of when the model was last created.
String arn
The model ARN.
ExternalModelSummary externalModel
The Amazon SageMaker model.
Map<K,V> outputs
The fraud prediction scores from Amazon SageMaker model.
String value
A statement containing a resource property and a value to specify filter condition.
String eventTypeName
Name of event type for which to get the deletion status.
String detectorId
The detector ID.
String detectorVersionId
The detector version ID.
String description
The detector version description.
List<E> externalModelEndpoints
The Amazon SageMaker model endpoints included in the detector version.
List<E> modelVersions
The model versions included in the detector version.
List<E> rules
The rules included in the detector version.
String status
The status of the detector version.
String lastUpdatedTime
The timestamp when the detector version was last updated.
String createdTime
The timestamp when the detector version was created.
String ruleExecutionMode
The execution mode of the rule in the dectector
FIRST_MATCHED indicates that Amazon Fraud Detector evaluates rules sequentially, first to last,
stopping at the first matched rule. Amazon Fraud dectector then provides the outcomes for that single rule.
ALL_MATCHED indicates that Amazon Fraud Detector evaluates all rules and returns the outcomes for
all matched rules. You can define and edit the rule mode at the detector version level, when it is in draft
status.
String arn
The detector version ARN.
String eventId
The event ID.
String eventTypeName
The event type associated with the detector specified for the prediction.
String detectorId
The detector ID.
String detectorVersionId
The detector version ID.
String predictionTimestamp
The timestamp that defines when the prediction was generated. The timestamp must be specified using ISO 8601 standard in UTC.
We recommend calling
ListEventPredictions first, and using the predictionTimestamp value in the response to provide
an accurate prediction timestamp value.
String eventId
The event ID.
String eventTypeName
The event type associated with the detector specified for this prediction.
String entityId
The entity ID.
String entityType
The entity type.
String eventTimestamp
The timestamp for when the prediction was generated for the associated event ID.
String detectorId
The detector ID.
String detectorVersionId
The detector version ID.
String detectorVersionStatus
The status of the detector version.
List<E> eventVariables
A list of event variables that influenced the prediction scores.
List<E> rules
List of rules associated with the detector version that were used for evaluating variable values.
String ruleExecutionMode
The execution mode of the rule used for evaluating variable values.
List<E> outcomes
The outcomes of the matched rule, based on the rule execution mode.
List<E> evaluatedModelVersions
Model versions that were evaluated for generating predictions.
List<E> evaluatedExternalModels
External (Amazon SageMaker) models that were evaluated for generating predictions.
String predictionTimestamp
The timestamp that defines when the prediction was generated.
String detectorId
The detector ID.
String detectorVersionId
The detector version ID.
String eventId
The unique ID used to identify the event.
String eventTypeName
The event type associated with the detector specified for the prediction.
List<E> entities
The entity type (associated with the detector's event type) and specific entity ID representing who performed the event. If an entity id is not available, use "UNKNOWN."
String eventTimestamp
Timestamp that defines when the event under evaluation occurred. The timestamp must be specified using ISO 8601 standard in UTC.
Map<K,V> eventVariables
Names of the event type's variables you defined in Amazon Fraud Detector to represent data elements and their corresponding values for the event you are sending for evaluation.
You must provide at least one eventVariable
To ensure most accurate fraud prediction and to simplify your data preparation, Amazon Fraud Detector will replace all missing variables or values as follows:
For Amazon Fraud Detector trained models:
If a null value is provided explicitly for a variable or if a variable is missing, model will replace the null value or the missing variable (no variable name in the eventVariables map) with calculated default mean/medians for numeric variables and with special values for categorical variables.
For imported SageMaker models:
If a null value is provided explicitly for a variable, the model and rules will use “null” as the value. If a variable is not provided (no variable name in the eventVariables map), model and rules will use the default value that is provided for the variable.
Map<K,V> externalModelEndpointDataBlobs
The Amazon SageMaker model endpoint input data blobs.
List<E> modelScores
The model scores. Amazon Fraud Detector generates model scores between 0 and 1000, where 0 is low fraud risk and 1000 is high fraud risk. Model scores are directly related to the false positive rate (FPR). For example, a score of 600 corresponds to an estimated 10% false positive rate whereas a score of 900 corresponds to an estimated 2% false positive rate.
List<E> ruleResults
The results from the rules.
List<E> externalModelOutputs
The model scores for Amazon SageMaker models.
Event event
The details of the event.
KMSKey kmsKey
The KMS encryption key.
String modelId
The model ID.
String modelType
The model type.
String modelVersionNumber
The model version number.
String trainingDataSource
The training data source.
TrainingDataSchema trainingDataSchema
The training data schema.
ExternalEventsDetail externalEventsDetail
The details of the external events data used for training the model version. This will be populated if the
trainingDataSource is EXTERNAL_EVENTS
IngestedEventsDetail ingestedEventsDetail
The details of the ingested events data used for training the model version. This will be populated if the
trainingDataSource is INGESTED_EVENTS.
String status
The model version status.
Possible values are:
TRAINING_IN_PROGRESS
TRAINING_COMPLETE
ACTIVATE_REQUESTED
ACTIVATE_IN_PROGRESS
ACTIVE
INACTIVATE_REQUESTED
INACTIVATE_IN_PROGRESS
INACTIVE
ERROR
String arn
The model version ARN.
IngestedEventsTimeWindow ingestedEventsTimeWindow
The start and stop time of the ingested events.
Long numberOfEvents
The number of stored events.
Long eventDataSizeInBytes
The total size of the stored events.
String leastRecentEvent
The oldest stored event.
String mostRecentEvent
The newest stored event.
String lastUpdatedTime
Timestamp of when the stored event was last updated.
String kmsEncryptionKeyArn
The encryption key ARN.
Map<K,V> labelMapper
The label mapper maps the Amazon Fraud Detector supported model classification labels (FRAUD,
LEGIT) to the appropriate event type labels. For example, if "FRAUD" and "
LEGIT" are Amazon Fraud Detector supported labels, this mapper could be:
{"FRAUD" => ["0"], "LEGIT" => ["1"]} or {"FRAUD" => ["false"],
"LEGIT" => ["true"]} or {"FRAUD" => ["fraud", "abuse"],
"LEGIT" => ["legit", "safe"]}. The value part of the mapper is a list, because you may have
multiple label variants from your event type for a single Amazon Fraud Detector label.
String unlabeledEventsTreatment
The action to take for unlabeled events.
Use IGNORE if you want the unlabeled events to be ignored. This is recommended when the majority of
the events in the dataset are labeled.
Use FRAUD if you want to categorize all unlabeled events as “Fraud”. This is recommended when most
of the events in your dataset are fraudulent.
Use LEGIT if you want to categorize all unlabeled events as “Legit”. This is recommended when most
of the events in your dataset are legitimate.
Use AUTO if you want Amazon Fraud Detector to decide how to use the unlabeled data. This is
recommended when there is significant unlabeled events in the dataset.
By default, Amazon Fraud Detector ignores the unlabeled data.
FilterCondition eventId
The event ID.
FilterCondition eventType
The event type associated with the detector.
FilterCondition detectorId
The detector ID.
FilterCondition detectorVersionId
The detector version ID.
PredictionTimeRange predictionTimeRange
The time period for when the predictions were generated.
String nextToken
Identifies the next page of results to return. Use the token to make the call again to retrieve the next page. Keep all other arguments unchanged. Each pagination token expires after 24 hours.
Integer maxResults
The maximum number of predictions to return for the request.
List<E> eventPredictionSummaries
The summary of the past predictions.
String nextToken
Identifies the next page of results to return. Use the token to make the call again to retrieve the next page. Keep all other arguments unchanged. Each pagination token expires after 24 hours.
String variableName
The name of the variable.
String variableType
The type of variable.
Float variableImportance
The relative importance of the variable. For more information, see Model variable importance.
Float fpr
The false positive rate. This is the percentage of total legitimate events that are incorrectly predicted as fraud.
Float precision
The percentage of fraud events correctly predicted as fraudulent as compared to all events predicted as fraudulent.
Float tpr
The true positive rate. This is the percentage of total fraud the model detects. Also known as capture rate.
Float threshold
The model threshold that specifies an acceptable fraud capture rate. For example, a threshold of 500 means any model score 500 or above is labeled as fraud.
String modelId
The model ID.
String modelType
The model type.
String description
The model description.
String eventTypeName
The name of the event type.
String createdTime
Timestamp of when the model was created.
String lastUpdatedTime
Timestamp of last time the model was updated.
String arn
The ARN of the model.
ByteBuffer byteBuffer
The byte buffer of the Amazon SageMaker model endpoint input data blob.
String contentType
The content type of the Amazon SageMaker model endpoint input data blob.
String eventTypeName
The event type name.
String format
The format of the model input configuration. The format differs depending on if it is passed through to SageMaker or constructed by Amazon Fraud Detector.
Boolean useEventVariables
The event variables.
String jsonInputTemplate
Template for constructing the JSON input-data sent to SageMaker. At event-evaluation, the placeholders for variable names in the template will be replaced with the variable values before being sent to SageMaker.
String csvInputTemplate
Template for constructing the CSV input-data sent to SageMaker. At event-evaluation, the placeholders for variable-names in the template will be replaced with the variable values before being sent to SageMaker.
String format
The format of the model output configuration.
Map<K,V> jsonKeyToVariableMap
A map of JSON keys in response from SageMaker to the Amazon Fraud Detector variables.
Map<K,V> csvIndexToVariableMap
A map of CSV index values in the SageMaker response to the Amazon Fraud Detector variables.
ModelVersion modelVersion
The model version.
Map<K,V> scores
The model's fraud prediction scores.
String modelId
The model ID.
String modelType
The model type.
String modelVersionNumber
The model version number.
String status
The status of the model version.
String trainingDataSource
The model version training data source.
TrainingDataSchema trainingDataSchema
The training data schema.
ExternalEventsDetail externalEventsDetail
The external events data details. This will be populated if the trainingDataSource for the model
version is specified as EXTERNAL_EVENTS.
IngestedEventsDetail ingestedEventsDetail
The ingested events data details. This will be populated if the trainingDataSource for the model
version is specified as INGESTED_EVENTS.
TrainingResult trainingResult
The training results.
String lastUpdatedTime
The timestamp when the model was last updated.
String createdTime
The timestamp when the model was created.
String arn
The model version ARN.
TrainingResultV2 trainingResultV2
The training result details. The details include the relative importance of the variables.
String outputVariableName
The output variable name.
String evaluationScore
The evaluation score generated for the model version.
PredictionExplanations predictionExplanations
The prediction explanations generated for the model version.
Float fpr
The false positive rate. This is the percentage of total legitimate events that are incorrectly predicted as fraud.
Float precision
The percentage of fraud events correctly predicted as fraudulent as compared to all events predicted as fraudulent.
Float tpr
The true positive rate. This is the percentage of total fraud the model detects. Also known as capture rate.
Float threshold
The model threshold that specifies an acceptable fraud capture rate. For example, a threshold of 500 means any model score 500 or above is labeled as fraud.
Float auc
The area under the curve (auc). This summarizes the total positive rate (tpr) and false positive rate (FPR) across all possible model score thresholds.
UncertaintyRange uncertaintyRange
Indicates the range of area under curve (auc) expected from the OFI model. A range greater than 0.1 indicates higher model uncertainity.
List<E> metricDataPoints
The model's performance metrics data points.
OFIModelPerformance modelPerformance
The model's overall performance score.
List<E> variableImpactExplanations
The details of the event variable's impact on the prediction score.
List<E> aggregatedVariablesImpactExplanations
The details of the aggregated variables impact on the prediction score.
Account Takeover Insights (ATI) model uses event variables from the login data you provide to continuously
calculate a set of variables (aggregated variables) based on historical events. For example, your ATI model might
calculate the number of times an user has logged in using the same IP address. In this case, event variables used
to derive the aggregated variables are IP address and user.
String name
The name.
String description
The description of the event type.
List<E> eventVariables
The event type variables.
List<E> labels
The event type labels.
List<E> entityTypes
The entity type for the event type. Example entity types: customer, merchant, account.
String eventIngestion
Specifies if ingestion is enabled or disabled.
List<E> tags
A collection of key and value pairs.
EventOrchestration eventOrchestration
Enables or disables event orchestration. If enabled, you can send event predictions to select AWS services for downstream processing of the events.
String modelEndpoint
The model endpoints name.
String modelSource
The source of the model.
String invokeModelEndpointRoleArn
The IAM role used to invoke the model endpoint.
ModelInputConfiguration inputConfiguration
The model endpoint input configuration.
ModelOutputConfiguration outputConfiguration
The model endpoint output configuration.
String modelEndpointStatus
The model endpoint’s status in Amazon Fraud Detector.
List<E> tags
A collection of key and value pairs.
String kmsEncryptionKeyArn
The KMS encryption key ARN.
The KMS key must be single-Region key. Amazon Fraud Detector does not support multi-Region KMS key.
String ruleId
The rule ID.
String description
The rule description.
String detectorId
The detector for which the rule is associated.
String ruleVersion
The rule version.
String expression
The rule expression.
String language
The rule language.
List<E> outcomes
The rule outcomes.
String lastUpdatedTime
Timestamp of the last time the rule was updated.
String createdTime
The timestamp of when the rule was created.
String arn
The rule ARN.
String eventId
The event ID to upload.
String eventTypeName
The event type name of the event.
String eventTimestamp
The timestamp that defines when the event under evaluation occurred. The timestamp must be specified using ISO 8601 standard in UTC.
Map<K,V> eventVariables
Names of the event type's variables you defined in Amazon Fraud Detector to represent data elements and their corresponding values for the event you are sending for evaluation.
String assignedLabel
The label to associate with the event. Required if specifying labelTimestamp.
String labelTimestamp
The timestamp associated with the label. Required if specifying assignedLabel.
List<E> entities
An array of entities.
Float fpr
The false positive rate. This is the percentage of total legitimate events that are incorrectly predicted as fraud.
Float precision
The percentage of fraud events correctly predicted as fraudulent as compared to all events predicted as fraudulent.
Float tpr
The true positive rate. This is the percentage of total fraud the model detects. Also known as capture rate.
Float threshold
The model threshold that specifies an acceptable fraud capture rate. For example, a threshold of 500 means any model score 500 or above is labeled as fraud.
Float auc
The area under the curve (auc). This summarizes the total positive rate (tpr) and false positive rate (FPR) across all possible model score thresholds.
UncertaintyRange uncertaintyRange
Indicates the range of area under curve (auc) expected from the TFI model. A range greater than 0.1 indicates higher model uncertainity.
List<E> metricDataPoints
The model's performance metrics data points.
TFIModelPerformance modelPerformance
The model performance score.
List<E> modelVariables
The training data schema variables.
LabelSchema labelSchema
Float auc
The area under the curve. This summarizes true positive rate (TPR) and false positive rate (FPR) across all possible model score thresholds. A model with no predictive power has an AUC of 0.5, whereas a perfect model has a score of 1.0.
List<E> metricDataPoints
The data points details.
OFITrainingMetricsValue ofi
The Online Fraud Insights (OFI) model training metric details.
TFITrainingMetricsValue tfi
The Transaction Fraud Insights (TFI) model training metric details.
ATITrainingMetricsValue ati
The Account Takeover Insights (ATI) model training metric details.
DataValidationMetrics dataValidationMetrics
The validation metrics.
TrainingMetrics trainingMetrics
The training metric details.
VariableImportanceMetrics variableImportanceMetrics
The variable importance metrics.
DataValidationMetrics dataValidationMetrics
TrainingMetricsV2 trainingMetricsV2
The training metric details.
VariableImportanceMetrics variableImportanceMetrics
AggregatedVariablesImportanceMetrics aggregatedVariablesImportanceMetrics
The variable importance metrics of the aggregated variables.
Account Takeover Insights (ATI) model uses event variables from the login data you provide to continuously
calculate a set of variables (aggregated variables) based on historical events. For example, your ATI model might
calculate the number of times an user has logged in using the same IP address. In this case, event variables used
to derive the aggregated variables are IP address and user.
String detectorId
The parent detector ID for the detector version you want to update.
String detectorVersionId
The detector version ID.
List<E> externalModelEndpoints
The Amazon SageMaker model endpoints to include in the detector version.
List<E> rules
The rules to include in the detector version.
String description
The detector version description.
List<E> modelVersions
The model versions to include in the detector version.
String ruleExecutionMode
The rule execution mode to add to the detector.
If you specify FIRST_MATCHED, Amazon Fraud Detector evaluates rules sequentially, first to last,
stopping at the first matched rule. Amazon Fraud dectector then provides the outcomes for that single rule.
If you specifiy ALL_MATCHED, Amazon Fraud Detector evaluates all rules and returns the outcomes for
all matched rules. You can define and edit the rule mode at the detector version level, when it is in draft
status.
The default behavior is FIRST_MATCHED.
String eventId
The ID of the event associated with the label to update.
String eventTypeName
The event type of the event associated with the label to update.
String assignedLabel
The new label to assign to the event.
String labelTimestamp
The timestamp associated with the label. The timestamp must be specified using ISO 8601 standard in UTC.
String name
The name of the list to update.
List<E> elements
One or more list elements to add or replace. If you are providing the elements, make sure to specify the
updateMode to use.
If you are deleting all elements from the list, use REPLACE for the updateMode and
provide an empty list (0 elements).
String description
The new description.
String updateMode
The update mode (type).
Use APPEND if you are adding elements to the list.
Use REPLACE if you replacing existing elements in the list.
Use REMOVE if you are removing elements from the list.
String variableType
The variable type you want to assign to the list.
You cannot update a variable type of a list that already has a variable type assigned to it. You can assign a variable type to a list only if the list does not already have a variable type.
String modelId
The model ID.
String modelType
The model type.
String majorVersionNumber
The major version number.
ExternalEventsDetail externalEventsDetail
The details of the external events data used for training the model version. Required if
trainingDataSource is EXTERNAL_EVENTS.
IngestedEventsDetail ingestedEventsDetail
The details of the ingested event used for training the model version. Required if your
trainingDataSource is INGESTED_EVENTS.
List<E> tags
A collection of key and value pairs.
Rule rule
The new rule version that was created.
String name
The name of the variable.
String defaultValue
The new default value of the variable.
String description
The new description.
String variableType
The variable type. For more information see Variable types.
String name
The name of the variable.
String dataType
The data type of the variable. For more information see Variable types.
String dataSource
The data source of the variable.
String defaultValue
The default value of the variable.
String description
The description of the variable.
String variableType
The variable type of the variable.
Valid Values:
AUTH_CODE | AVS | BILLING_ADDRESS_L1 | BILLING_ADDRESS_L2 | BILLING_CITY | BILLING_COUNTRY | BILLING_NAME | BILLING_PHONE | BILLING_STATE | BILLING_ZIP | CARD_BIN | CATEGORICAL | CURRENCY_CODE | EMAIL_ADDRESS | FINGERPRINT | FRAUD_LABEL | FREE_FORM_TEXT | IP_ADDRESS | NUMERIC | ORDER_ID | PAYMENT_TYPE | PHONE_NUMBER | PRICE | PRODUCT_CATEGORY | SHIPPING_ADDRESS_L1 | SHIPPING_ADDRESS_L2 | SHIPPING_CITY | SHIPPING_COUNTRY | SHIPPING_NAME | SHIPPING_PHONE | SHIPPING_STATE | SHIPPING_ZIP | USERAGENT
String lastUpdatedTime
The time when variable was last updated.
String createdTime
The time when the variable was created.
String arn
The ARN of the variable.
String name
The name of the variable.
String dataType
The data type of the variable.
String dataSource
The data source of the variable.
String defaultValue
The default value of the variable.
String description
The description of the variable.
String variableType
The type of the variable. For more information see Variable types.
Valid Values:
AUTH_CODE | AVS | BILLING_ADDRESS_L1 | BILLING_ADDRESS_L2 | BILLING_CITY | BILLING_COUNTRY | BILLING_NAME | BILLING_PHONE | BILLING_STATE | BILLING_ZIP | CARD_BIN | CATEGORICAL | CURRENCY_CODE | EMAIL_ADDRESS | FINGERPRINT | FRAUD_LABEL | FREE_FORM_TEXT | IP_ADDRESS | NUMERIC | ORDER_ID | PAYMENT_TYPE | PHONE_NUMBER | PRICE | PRODUCT_CATEGORY | SHIPPING_ADDRESS_L1 | SHIPPING_ADDRESS_L2 | SHIPPING_CITY | SHIPPING_COUNTRY | SHIPPING_NAME | SHIPPING_PHONE | SHIPPING_STATE | SHIPPING_ZIP | USERAGENT
String eventVariableName
The event variable name.
String relativeImpact
The event variable's relative impact in terms of magnitude on the prediction scores. The relative impact values consist of a numerical rating (0-5, 5 being the highest) and direction (increased/decreased) impact of the fraud risk.
Float logOddsImpact
The raw, uninterpreted value represented as log-odds of the fraud. These values are usually between -10 to +10, but range from - infinity to + infinity.
A positive value indicates that the variable drove the risk score up.
A negative value indicates that the variable drove the risk score down.
Copyright © 2025. All rights reserved.