PredictionModelStatus Type Definition

[This topic is pre-release documentation and is subject to change.]

Prediction Model Status represents the status of the machine learning training model used for predictive analysis.

PredictionModelStatus Properties

Property names appended with asterisks* are required.

Property JSON type Description
tenantId string Name of the owning hub (readonly)
predictionName string Name of the associated prediction (readonly)
predictionGuidId string (GUID) ID of the associated prediction (readonly)
status* string Current point in prediction model life cycle: [New | Provisioning | ProvisioningFailed | PendingDiscovering | Discovering | PendingFeaturing | Featuring | FeaturingFailed | PendingTraining | Training | TrainingFailed | Evaluating | EvaluatingFailed | PendingModelConfirmation | Active | Deleted | HumanIntervention | Failed]
message string Model status message (readonly)
trainingSetCount integer Count of the test set (readonly)
testSetCount integer Count of the test set (readonly)
validationSetCount integer Count of the validation set (readonly)
trainingAccuracy integer Training accuracy (readonly)
signalsUsed integer Signals used in the training (readonly)
modelVersion string Version of the model (readonly)