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. Is it possible to alter an Azure ML Gallery submission after it has been deposited.

    Is it possible to alter an Azure ML Gallery submission after it has been deposited.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  2. Rethink the 8 minute trial

    The 8 minute trial is so restrictive you needn't bother with it.
    The 8 minute countdown is so daunting it almost dissuaded me staying on the site.

    You can't even upload your own datasets?

    I understand that some of this is to stop people from gaming the system and getting lots of free machine learning compute time but you're putting so many people off in the process.

    Oh and the tutorial is really really bad.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  3. Improve the tutorial

    The tutorial needs improving. It seems to draw big assumptions about the users knowledge of machine learning.

    The aim of the experiment needs clarifying before the drag and drop starts happening.

    The choice of algorithm needs explaining!?

    Every step needs explaining more.

    The conclusion screen doesn't clearly explain what the experiment proved.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  4. Add more languages as sample Request Response API

    My students would like to get some more examples in different languages at the bottom of the sample request response API. Is there any chance for PHP, Java, Scala?

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  5. [Bug] Issue with UI on Azure Machine Learning

    When pushing a model to the gallery on the Machine Learning UI, you can't close the window because the z-index is lower than the navigation bar's and it's blocking the close button.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  6. Azure ML : possbility to save custom R script modules to library and reuse across multiple experiments

    Azure ML : possbility to save custom R script modules to library and reuse them across multiple experiments.

    6 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  7. Add Package ‘DMwR’

    I need to use lofactor function. This function obtain local outlier factors using the LOF algorithm.

    90 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  8. r script

    I have a R-Script produced under R version 3.2.1 , imported the script within Azure ML Studio but facing issue in R-script execution. It came to notice that current supported R Version is 3.1.0 in Azure ML. Some of the packages used in script are not available in Azure ML. Hence, I am passing the packages (incl dependencies) compatible with R version 3.1.0 in Azure ML with zipped dataset. But some of the dependencies (compatible with 3.1.0) are not available from CRAN archive, nor from GitHub e.g “dplyr”. Is there any way out?
    Consumed R-Packages:
    • qdap

    • SnowballC
    •…

    14 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  9. Add a folder structure to the my experiments and webservices tabs

    Please add the ability to create sub-folders and move around items in the 'My experiments' and 'webservices' tabs, to help us organise our workspaces better.

    30 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  10. Specify features that should be ignored by a model

    I have a dataset with features that should be ignored when training/applying a model, but which I want to keep to facilitate interpretation of the results. For example, IDs, dates, name of the data source, etc.
    I find that in some cases, the trained models consider these features as significant (for example they have a high weight in a regression model), even though their importance is a coincidence. Therefore, there should be a way to specify which attributes should be excluded from a model's inputs.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  11. Add error message for failed experiments

    When I run an experiment, I get an error failed without any additional message, even for a super simple step. I can see absolutely no details about the error. The error is not described on rollover or in the run history either.

    Attached: screenshot and dataset

    8 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  12. Saving Graphic R output files through a C# API or save in azure blobs

    Feature to save the Execute R script graphic file or through C# API or to save it in azure blobs

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  13. Module for feature ranking

    Provide a module for feature ranking. So, for a model built with a set of features, we should be able rank the features in order of relevance.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  14. Add R plugin in Visual Studio for syntax highlight and turns VS in R IDE

    The plugin exist for R Revolution, just add it like Python plugin for V IDE.

    39 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  15. Deployment of ML experiments programmatically

    Presently, the ML workspace and experiments needs to be manually created. There is no capability to create new workspace and experiments using some scripting like, C# or PowerShell. It will be good, if we can take backup or generate deployable files from existing ML experiments which in turn can be used for deploying it to another/same subscription with same/different name.

    94 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    2 comments  ·  Flag idea as inappropriate…  ·  Admin →
  16. Show variable importance after experiment runs.

    One of the more interesting things to observe when building a model is seeing what variables were determined to be most important.

    Allowing us to see various variable weights (p values for linear regression, importance in tree based models, etc.) would make experiments more insightful.

    59 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  17. Randomized Split is not giving apt split

    Even with Seed value as 0, Randomized split is giving me same training set without any difference not a single one. As it take seed from system time, so it was run in different time to test the same. It is actaually behaiving in same way as it is working with some seed.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  18. Experiment Name Missing First Character

    Please fix the Experiment Name feature. The first character of experiment names is being lost sometimes. Seems like a bug in the UI.

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  19. 28 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  20. Order of columns when scoring a model

    If a dataset with many columns is used the "viewer" is truncating after x columns. I assume this is to keep things running smooth, however, would it be possible to by default not to trim informative columns like the outcome of a scored model? That way you can quickly use the web GUI to see how well your model is doing without having to use another "project columns".

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    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