Skip to main content

Data Retention and Purging

The Machine Learning service builds upon the to provide automated machine learning functionality to customers. The application leverages the application security work of the underlying platform and extends the same data security strategies within application-specific areas.

Data Retention and Purging: The Data Lifecycle

Machine Learning allows you to import a dataset into the application for analysis and predictive modeling. The server copies imported datasets and then analyzes the data to provide insights to you. Projects maintain both imported data and derived data in cloud storage accessible only to the project owner. You have the option to import updated datasets and rerun analysis and modeling with that updated data. This creates a history of linked objects associated with a project.

Projects have a lifecycle from creation to deletion. When you delete a project, Machine Learning marks both imported and derived data for deletion. The deletion occurs asynchronously and might take several minutes based on the number of objects in the history of linked objects.