none
Change Datasets on a model by code

    Question

  • Hi,

    I have a model on Azure ML Studio and I want to replace datasets on my model everyday automatically so is it workable ? Should I use the notebook ? How can I run the notebook everyday automatically ?

    Thank you

    Monday, January 7, 2019 9:29 AM

Answers

  • Hi, 

    Like stated by FelypeBastos, you could store your data in Azure Blob, then import it using 'Import Data' module: https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/import-from-azure-blob-storage

    Then set the module to have a web service parameter as stated in this document: https://docs.microsoft.com/en-us/azure/machine-learning/studio/web-service-parameters

    One option would be to deploy your experiment as a web service and score it daily using Azure Data Factory for scheduling: https://docs.microsoft.com/en-us/azure/data-factory/transform-data-using-machine-learning

    Regards,
    Jaya

    Monday, January 7, 2019 8:23 PM

All replies

  • Hey Alex, try Import Data on experiment.

    When you call the web service, the dataset will be loaded from "Import Data".

    For update the dataset automatically, you will need to create a routine to update the dataset.

    Monday, January 7, 2019 5:11 PM
  • Hi, 

    Like stated by FelypeBastos, you could store your data in Azure Blob, then import it using 'Import Data' module: https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/import-from-azure-blob-storage

    Then set the module to have a web service parameter as stated in this document: https://docs.microsoft.com/en-us/azure/machine-learning/studio/web-service-parameters

    One option would be to deploy your experiment as a web service and score it daily using Azure Data Factory for scheduling: https://docs.microsoft.com/en-us/azure/data-factory/transform-data-using-machine-learning

    Regards,
    Jaya

    Monday, January 7, 2019 8:23 PM