question

LauriLehman-8626 avatar image
1 Vote"
LauriLehman-8626 asked ·

Cross validation with Azure ML Python SDK

Does Azure ML Python SDK support cross validation? In the graphical Designer, there is a cross validation module but I haven't found anything similar in the SDK documentation.

Of course, there probably exist many frameworks for cross-validation in Python but it would be nice to have a native Azure ML module for cross-validation, similar to the Hyperparameter tuning module. I would like to be able to give a single Dataset object and the Azure ML framework would take care of slicing the source Dataset into separate fold datasets. Azure ML should also take care of assigning different folds to the compute nodes and running the folds in parallel.

azure-machine-learning
· 3
10 |1000 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

Can you please confirm are you expecting this in Azure AutoML? If yes, AutoML supports cross-validation.

0 Votes 0 ·

I was thinking custom ML models (trained as Estimators, for example). I believe AutoML already has this feature.

1 Vote 1 ·

Okay, We have forwarded to product team to check, will update on the same. Yes AutoML supports internally split or cross-validation. In the future, we're planning for AutoML to also support a Test dataset, so you could provide the following as parameters to the AutoMLConfig class:

Train Dataset,
Validation Dataset,
Test Dataset


1 Vote 1 ·

1 Answer

Ram-msft avatar image
1 Vote"
Ram-msft answered ·

@LauriLehman-8626 It's not supported to use Designer built-in module in python SDK today but we are working on the module SDK private preview which can enable this. We would like understand more about your use-case, can you please send an email to Azcommunity@microsoft.com so that we can invite you to the private preview if you are interested.


·
10 |1000 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.