Quantify conformity or compliance between different datasets

Anonymous
2022-01-03T16:05:21.097+00:00

Hello,
I have generated regression model using Azure design for a specific dataset of which I know the value to be predicted so I can evaluate the model performance and tune hyper-parameters. But now I would like to apply this trained model to a new and different dataset of which I do not know the values to be predicted during the regression so I cannot quantify the performance of the model for this testing dataset. However, for doing so I would like to see if the testing dataset is representative or similar to the trained dataset to evaluate if the prediction will be accurate. Is there a way in Azure to measure the conformity or compliance of a testing dataset to be similar or comparable to the training dataset?
Thank you!

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
2,691 questions
Azure Blob Storage
Azure Blob Storage
An Azure service that stores unstructured data in the cloud as blobs.
2,579 questions
Azure Data Explorer
Azure Data Explorer
An Azure data analytics service for real-time analysis on large volumes of data streaming from sources including applications, websites, and internet of things devices.
503 questions
0 comments No comments
{count} votes

1 answer

Sort by: Most helpful
  1. GiftA-MSFT 11,161 Reputation points
    2022-01-03T21:26:01.087+00:00

    Hi, with dataset monitor, you can detect data drift on datasets. Please review the documentation and let us know if it helps your scenario.


    --- *Kindly Accept Answer if the information helps. Thanks.*

    1 person found this answer helpful.
    0 comments No comments