Update: Microsoft will be moving away from UserVoice sites on a product-by-product basis throughout the 2021 calendar year. We will leverage 1st party solutions for customer feedback. Learn more here.

Machine Learning

Welcome! The Azure Machine Learning team invites you to share and vote for features to help you build, manage, and deploy custom machine learning models.

Have a technical question, or want to learn more? Please visit our documentation, MSDN forum or StackOverflow.


  1. Make it like Orange

    Can you actively analyze data in a new way and have the models update and the visualizations update instantly to determine if the new analysis is a good idea?
    Best feature of Orange is instant feedback on your progress, will you have it?

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  2. Expose Cluster Purpose On Arm Template Deployments

    In the AzureML SDK you are able to specify when attaching an AKS Compute Instance to the workspace whether the purpose is DEVTEST or FASTPROD.

    For development environments you can specify DEV_TEST so that the minimum 3 Nodes are not imposed on the cluster (keep cost down on test environmets).

    I noticed that this functionality does not seemed to be exposed when using ARM Templates (https://docs.microsoft.com/en-us/azure/templates/microsoft.machinelearningservices/2020-01-01/workspaces/computes#AKS).

    This means that if you are trying to easily spin up test environments using ARM templates it will set the purpose to FAST_PROD which means that you will need a…

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  3. Add the function to delete any version of the dataset

    If you update a dataset with incorrect data, there is no way to delete that version. It would be helpful if there was a function that could delete only the specific version you want to delete.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  4. Add a chart for evaluating a time-series forecast model

    Evaluate automated machine learning experiment results (https://docs.microsoft.com/en-us/azure/machine-learning/how-to-understand-automated-ml)

    This site has charts and metrics for Classification and Regression, but no for time-series forecast models. Customers have asked us to add a chart like the one attached, so please consider it.

    "chart.png" file: a comparison chart of real data and prediction results.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  5. Make AutoML a module in the Designer

    It would be very useful if AutoML could be used from within the Designer.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  6. Read HHTP response in Azure Machine Learning

    I created a logic app that invoke a Machine Learning pipeline when the blob storage is changed. When I create the method HHTP request pass two parameters in the JSON: file store and the file path (the same example as: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-trigger-published-pipeline), however the Machine Learning receive the response but the file of pipelines is not change (the value is static). How can I read the HTTP response in Azure Machine Learning and pass this information in the pipline endpoint so that the result is always dynamic depending on blob storage?

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  7. Monetization of pre-trained ML Models via AI-store

    As a user it would be great to be able to use an existing pre-trained model for some of my AI use-cases. This model can be supplied by a third party supplier. As a user I pay a fee per minute of usage.
    The same approach is currently available witin the AWS Sagemaker marketplace.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  8. Support ParallelRunStep and Many Models in Visual Designer

    In order to implement a Many Models solution code is required[1]. This solution with the ParallelRunStep should be accomplishable via the Visual Designer.

    I notice that an icon is available for this step AFTER it's created via code, but not an option in the pipeline editor.

    [1] <https://github.com/microsoft/solution-accelerator-many-models>

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  9. Resync Keys

    It is very said we need manually "Resync Keys" in ML studio. We can not even find a API to call resync key in power-shell.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →

    Hi Azure Customer,

    Thank you for your feedback here. Since Azure Machine Learning Studio will be deprecated. We will not be supporting in Studio, investigtaing how Resync Keys works with visual interface

    We will get back to you soon once we have any update.

    Thank you for your understanding.

    Regards,
    Azure CXP Community

  10. Need a module to identify Low / High Correlated feature with Target Variable as well as High Correlated input features.

    A new module which outputs High/low Correlated input features with Target Variable as well as to identify High correlated input features.

    This module would help to create better Trained Data Set and that would turn to create better ML model.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  11. Allow Properties Shared by Multiple Items to be Edited In a Single Place

    If I select several items in an Azure ML Studio experiement that all share a common property, I should be able to set that property in one place - once - and not have to edit each item separately.

    A huge portion of what data scientist are doing in an experiment is trying out different classification algorithms, training them, scoring them, and evaluating and comparing results. This makes for a lot of repetitive, monotonous work. And it means we have to do exactly the same edits (like setting a dependent variable) across a bunch of different items.

    There's no reason…

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  12. How to integrate DevOps approach with VM Notebook in AMLS ?

    Today, the platform is amazingly evolving in term of integration. We can now monitor and manage almost everything in the visual interface.

    My question is how can we improve even more this experience by adding the DevOps part (instead of having to go threw Azure Notebooks, or Visual Studio) to the Notebook VM and Jupyter environment.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  13. Edit or remove experiments in Automated Machine Learning

    I just tried the Automated Machine Learning service with one of my older projects, simply out of curiosity.

    I created an experiment, with a compute setup, and then I wanted to add another compute instead. The only way it can be done is by creating a new experiment with the same name, but a different compute setup.

    Should it be possible to simply edit the experiment that was generated previously from the dashboard?

    I can see, that there is an idea of keeping the previous run results, but when there is a cancelled run, or I cannot benefit from it,…

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  14. Default compute target

    An awesome feature has been added to the new Visual interface: when I want to run my experiment, I have to specify what kind of machine I want to run it on.

    When the experiment has to run more than once, I found only one – tiny bit – annoying feature: every time you run the experiment, you have to choose a compute target. What if I want to use the same target for the next few days, or even longer?

    I wish there was a possibility to set a default size, and use that, unless I want to scale…

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  15. "az ml run show -r runid" pops surprised output

    When I am using command "az ml run show -r runid",
    This the output:

    az ml run show -r b4c190b4-fd00-43ca-bf31-639221a9485c
    Error, default experiment not set and experiment name parameter not provided.
    Please provide a value for the experiment name parameter.

    The problem is I can't provide experiment name, that's the information I want to query for...

    Could we refine this command? I can provide -w workspace instead.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  16. We need an example of using AutoML in Azure DevOps rather than a simple sklearn model in the Python MLOps example?

    Pelase Add good example in AzureDevOps using a video tutorial on your websiten rather than the MLOPs Python sample and how we can use it in Azure DevOps.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  17. 3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  18. define and run pipeline using Logic App

    It's super make sense this type of workflow via GUI, not only code. Please refere the following service as well.


    • Azure DevOps

    • Azure Data Factory

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  19. Enable Machine Learning Studio for Government Cloud

    Please enable azure machine learning services for government cloud customers

    Currently there is no timeline for releasing ML services to Gov Cloud users, but we are interested in the service.

    https://azure.microsoft.com/en-us/global-infrastructure/services/?products=machine-learning-studio

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  20. model_data_collection should log failed API request too

    We plan to leverage modeldatacollection message type in traces to log our API request payload (instead of printing the payload to STDOUT) only to realize that failed requests were not logged. It's actually more useful to log payload for failed requests so we can reproduce the issue rapidly.

    3 votes
    Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  • Don't see your idea?

Machine Learning

Categories

Feedback and Knowledge Base