none
How do I remove the label and other unnecessary attributes when I deploy my model? RRS feed

  • Question

  • I've been following the tutorial to deploy a credit risk model: https://docs.microsoft.com/en-us/azure/machine-learning/studio/tutorial-part3-credit-risk-deploy

    At the point of deployment, the docs state:

    One important thing to note is that if your original dataset contained the label, then the expected schema from the web input will also expect a column with the label! A way around this is to remove the label, and any other data that was in the training dataset, but will not be in the web inputs, before connecting the web input and training dataset into a common module.

    This is exactly why I'm doing the tutorial, to find out how to remove the label from the input schema but the paragraph above doesn't clearly tell me what to do, rather just to remove it. I can't find any information on this elsewhere. Can anyone explain how I would do this? I can't see why anyone would want to deploy a model that takes the predictor variable as an input. Assumed it would be removed automatically when the model is deployed.

    Ps I tried removing the predictor in the predictive experiment but it then fails when I try to run it because of this.

    Thanks,

    Luke

    Monday, July 8, 2019 4:08 PM

All replies

  • Hi,

    Please refer the following link to remove label column from the "Select Columns Dataset" from the predictive experiment as it will not be used in the prediction, but the Generated Predictive Experiment need to tweak a little bit to run the experiment successfully. I hope this helps!


    Tuesday, July 9, 2019 9:56 AM
    Moderator
  • Do you not think it is worth documenting this properly? Seems like a common task in ML Studio but Azure apparently assume that people will just automatically know how to do this when it's actually a bit of a fiddly procedure.

    Tuesday, August 6, 2019 2:42 PM
  • Hi, We have already forwarded this feedback to Azure Machine Learning Studio content team they are working on the document for this scenario to incorporate.

    Thanks





    Wednesday, August 7, 2019 6:13 AM
    Moderator