datadrift Package
Contains functionality to detect when model training data has drifted from its scoring data.
In machine learning, data drift is the change in model input data that leads to model performance degradation. It is one of the top reasons where model accuracy degrades over time, thus monitoring data drift helps detect model performance issues. This package enables you detect and alert on data drift.
The DataDriftDetector class enables you to configure a data monitor object which then can be run as a job to analyze data drift. Data drift jobs can be run interactively or enabled to run on a schedule. You can set up alerts when data drift exceeds a threshold with the AlertConfiguration class.
Modules
alert_configuration |
Contains functionality for configuring data drift alerts in Azure Machine Learning. |
datadriftdetector |
Contains core functionality for detecting data drift between two datasets in Azure Machine Learning. Data drift is measured through datasets or deployments, and relies on the Dataset API. |
Classes
AlertConfiguration |
Represents alert configuration for data drift jobs. The AlertConfiguration class allows for setting configurable alerts (such as email) on DataDriftDetector jobs. Alert configuration can be specified when using one of the create methods of the DataDriftDetector class. Constructor. Allows for setting configurable alerts (such as email) on DataDriftDetector jobs. |
DataDriftDetector |
Defines a data drift monitor that can be used to run data drift jobs in Azure Machine Learning. The DataDriftDetector class enables you to identify drift between a given baseline and target dataset. A DataDriftDetector object is created in a workspace by either specifying the baseline and target datasets directly. For more information, see https://aka.ms/datadrift. Datadriftdetector constructor. The DataDriftDetector constructor is used to retrieve a cloud representation of a DataDriftDetector object associated with the provided workspace. |
Metric |
Represents a metric returned in a data drift analysis. The Metric class is for internal usage only. Use the get_output method of a DataDriftDetector object to return metrics. Metric constructor. |
ModelServingDataset |
Represents a dataset used internally when a model-based DataDriftDetector object is created. A model-based DataDriftDetector enables you to calculate data drift between a model's training dataset and its scoring dataset. To create a model-based DataDriftDetector, use the <xref:azureml.datadrift.DataDriftDetector.create_from_model> method. Constructor. |
Enums
MetricType |
Defines types of metrics returned in a data drift analysis. Use the get_output method of a DataDriftDetector object to return metrics. |