Machine Learning

Welcome! You can use this site to tell the Azure Machine Learning team what features you would like to see.
Azure Machine Learning enables developers and data scientists to build advanced analytic web services in minutes and hours, eliminating the heavy lifting associated with deploying machine learning in modern data driven applications.

How can we improve Microsoft Azure Machine Learning?

You've used all your votes and won't be able to post a new idea, but you can still search and comment on existing ideas.

There are two ways to get more votes:

  • When an admin closes an idea you've voted on, you'll get your votes back from that idea.
  • You can remove your votes from an open idea you support.
  • To see ideas you have already voted on, select the "My feedback" filter and select "My open ideas".
(thinking…)

Enter your idea and we'll search to see if someone has already suggested it.

If a similar idea already exists, you can support and comment on it.

If it doesn't exist, you can post your idea so others can support it.

Enter your idea and we'll search to see if someone has already suggested it.

  1. Option to allow training for more than 24 hours

    The utilities available in ML Studio is really convenient for me, however my dataset (around 1m rows) is taking too long to train, and I am only using a 2 layer FC NN. I think most of us won't mind paying for it.

    1 vote
    Vote
    Sign in
    Check!
    (thinking…)
    Reset
    or sign in with
    • facebook
    • google
      Password icon
      Signed in as (Sign out)
      You have left! (?) (thinking…)
      2 comments  ·  Flag idea as inappropriate…  ·  Admin →

      Hi, When I responded. I was under the impression that you were using AML services and not AML Studio. For AML services, there is an option to choose low priority VM. https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-set-up-training-targets

      I’m not familiar with AML Studio so I am going to leave this question open. If you are new to AML studio and the adoption to AML services, isn’t going to be too high, it maybe worth considering using AML services.

      Thank you and Happy New Year!

    • Model data collection for ACI deployed models

      As far as I can tell from the documentation it is only possible to enable data collection on models deployed AKS. In scenarios where it is not desirable to create and manage Kubernetes in a production environment it would be useful to have this feature also for ACI based deploys.

      1 vote
      Vote
      Sign in
      Check!
      (thinking…)
      Reset
      or sign in with
      • facebook
      • google
        Password icon
        Signed in as (Sign out)
        You have left! (?) (thinking…)
        1 comment  ·  Flag idea as inappropriate…  ·  Admin →
      • Improve the amount of info provided when Send fails in the Test in Web Chat blade for Cognitive Services chat bots

        Azure Portal "Test in Web Chat" blade for bots could be much more helpful. Have a "show errors" option that defaults to on.

        When an error occurs, instead of the cryptic "couldn't send" error, at least show the HTTP error, if any. Client knows whether bot is alive or not and maybe bad path or internal server error.

        If Application Insights is enabled for the bot's web app, dump the AI data for the request to the web page, or provide a link where it can be downloaded, instead of me having to dig for the specific AI data.

        1 vote
        Vote
        Sign in
        Check!
        (thinking…)
        Reset
        or sign in with
        • facebook
        • google
          Password icon
          Signed in as (Sign out)
          You have left! (?) (thinking…)
          1 comment  ·  Flag idea as inappropriate…  ·  Admin →
        • Please add function that can install "CLI extension for Azure Machine Learning service" at Docker Azure CLI

          Please add function that can install "CLI extension for Azure Machine Learning service" at Docker Azure CLI

          4 votes
          Vote
          Sign in
          Check!
          (thinking…)
          Reset
          or sign in with
          • facebook
          • google
            Password icon
            Signed in as (Sign out)
            You have left! (?) (thinking…)
            0 comments  ·  Flag idea as inappropriate…  ·  Admin →
          • Ability to run experiments without using SSH port

            A lot of enterprise customers have concerns opening SSH port even with VNet enabled and using Azure ML service tag.

            We should find a way to run experiments on computes like DSVM without opening up access to SSH port.

            4 votes
            Vote
            Sign in
            Check!
            (thinking…)
            Reset
            or sign in with
            • facebook
            • google
              Password icon
              Signed in as (Sign out)
              You have left! (?) (thinking…)
              0 comments  ·  Flag idea as inappropriate…  ·  Admin →
            • 1 vote
              Vote
              Sign in
              Check!
              (thinking…)
              Reset
              or sign in with
              • facebook
              • google
                Password icon
                Signed in as (Sign out)
                You have left! (?) (thinking…)
                1 comment  ·  Flag idea as inappropriate…  ·  Admin →
              • is this forum still alive?

                Any roadmap for ML Studio?
                I see very old posts...

                1 vote
                Vote
                Sign in
                Check!
                (thinking…)
                Reset
                or sign in with
                • facebook
                • google
                  Password icon
                  Signed in as (Sign out)
                  You have left! (?) (thinking…)
                  1 comment  ·  Flag idea as inappropriate…  ·  Admin →
                • Mark models as faulty

                  We would like to manually mark some models as faulty so that they cannot be used or restored.

                  1 vote
                  Vote
                  Sign in
                  Check!
                  (thinking…)
                  Reset
                  or sign in with
                  • facebook
                  • google
                    Password icon
                    Signed in as (Sign out)
                    You have left! (?) (thinking…)
                    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
                  • Luis - Add an Azure Maps key property to LuisPredictionOptions - and have Luis call Azure Maps

                    First of all, please add a feedback category for Luis to this site.

                    The LuisPredictionOptions class allows the caller to provide a Bing key, so Luis can perform spell-checking.

                    Please add a similar property to LuisPredictionOptions for Azure Maps key.

                    Luis's recognition of US cities and states is unpredictable and incomplete. The geographyV2 entity finds some cities and not others.

                    Please allow the caller provide an Azure Maps key, so Luis can recognize all cities, POI's, zip codes, etc. rather than each dev having to reinvent this wheel and make multiple calls to Azure to perform entity identification.

                    3 votes
                    Vote
                    Sign in
                    Check!
                    (thinking…)
                    Reset
                    or sign in with
                    • facebook
                    • google
                      Password icon
                      Signed in as (Sign out)
                      You have left! (?) (thinking…)
                      0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                    • return more specific info for geographyV2

                      Using Luis, I have an utterance "food near boston ma". I have a pattern "food near [geographyV2]" and one "food near [geographyV2] [geographyV2]". Luis finds the "boston" entity but not "ma". Training and assigning doesn't help. Luis should recognize state abbreviations. Luis could recognize "boston, ma" as one geographical entity and return an object which specifies the city unambiguously. If the utterance is "springfield, ma" Luis would recognize "Springfield" as a city, but how do I know which Springfield if all I have is a text string with that single word? Then my code has to look for the "ma"…

                      1 vote
                      Vote
                      Sign in
                      Check!
                      (thinking…)
                      Reset
                      or sign in with
                      • facebook
                      • google
                        Password icon
                        Signed in as (Sign out)
                        You have left! (?) (thinking…)
                        0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                      • Please upgrade the version of R to R 3.5.2 used in azure

                        I run R codes in azure and according to my understanding it does not support the functions used in R 3.5.2 since azure uses an older version of R. Would be greatly helpful if you can upgrade the version to R 3.5.2

                        1 vote
                        Vote
                        Sign in
                        Check!
                        (thinking…)
                        Reset
                        or sign in with
                        • facebook
                        • google
                          Password icon
                          Signed in as (Sign out)
                          You have left! (?) (thinking…)
                          0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                        • Make getting an ID for Cognitive Services less mysterious

                          I wanted to try the Face API. I found:
                          https://westus.dev.cognitive.microsoft.com/docs/services/563879b61984550e40cbbe8d/operations/563879b61984550f30395236/console

                          But I needed the Ocp-Apim-Subscription-Key. You have a page that solves exactly my problem:
                          https://azure.microsoft.com/en-us/try/cognitive-services/?api=face-api

                          But it took me over ninety minutes of searching, including signing up for an account I don't need (yet...), and trying to follow outdated clues in the StackOverflow forum, to stumble onto the clue that led to that keys page!

                          My 142 page load struggle is shown in the attached file. Please post a prominent link to the "Try" page on as many of those as possible! I could have seen success in three minutes…

                          3 votes
                          Vote
                          Sign in
                          Check!
                          (thinking…)
                          Reset
                          or sign in with
                          • facebook
                          • google
                            Password icon
                            Signed in as (Sign out)
                            You have left! (?) (thinking…)
                            0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                          • 1:1 R packages support from Training to Web Service Deployment

                            I was successfully able to use the stringr package for my R module while the experiment was in Training.

                            However, when I attempted to deploy my experiment I was greeted with the "unable to install package stringr" when it was running for the first time in prediction mode.

                            Could we have some kind of documentation on which R packages are supported in Prediction (aka Production) versus Training (aka Development)?

                            Thank you.

                            1 vote
                            Vote
                            Sign in
                            Check!
                            (thinking…)
                            Reset
                            or sign in with
                            • facebook
                            • google
                              Password icon
                              Signed in as (Sign out)
                              You have left! (?) (thinking…)
                              0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                            • Role based access needed for Azure ML Service

                              We have multiple customers asking for the following role based scenarios,

                              - Experimentation: Define roles for users who can only submit the jobs, who can only manage computes etc.
                              - Inferencing: Define roles as explained below,

                              I’m trying to run a standard 3 env (build/test/live) set up with a single common AML Workspace (so that the immutable images tested are promoted, not re-created in a different Workspace instance).

                              I’d need to be able to achieve the below:
                              • Data Scientists can register models & create images within the AML Workspace
                              • Data Scientists can deploy images to ACI or Build…

                              1 vote
                              Vote
                              Sign in
                              Check!
                              (thinking…)
                              Reset
                              or sign in with
                              • facebook
                              • google
                                Password icon
                                Signed in as (Sign out)
                                You have left! (?) (thinking…)
                                0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                              • Add detailed model training progress information in Machine Learning Studio experiment view

                                At now, the only training progress indicator in experiment view is a time elapsed counter. The idea is to extend number and types of possible counters while the model is being trained. Such information could possibly be: current training iteration, progress percentage, etc.

                                6 votes
                                Vote
                                Sign in
                                Check!
                                (thinking…)
                                Reset
                                or sign in with
                                • facebook
                                • google
                                  Password icon
                                  Signed in as (Sign out)
                                  You have left! (?) (thinking…)
                                  0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                • Allow UPDATE a table in Export Data transformation

                                  Currently I can only INSERT rows to SQL table in my SQL Server in Azure using Export Data transformation.
                                  It would be useful to allow also UPDATE in cases when we need to update prediction scores of already "predicted" probabilities in the target table.
                                  Thanks

                                  1 vote
                                  Vote
                                  Sign in
                                  Check!
                                  (thinking…)
                                  Reset
                                  or sign in with
                                  • facebook
                                  • google
                                    Password icon
                                    Signed in as (Sign out)
                                    You have left! (?) (thinking…)
                                    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                  • More informative error message when building the docker image fails.

                                    I was trying to deploy a scoring pipeline on Azure using AML SDK in Python. The scoring pipeline needs fbprophet library to train the forecasting model on the fly, and forecast for the future. I added fbprophet into the pip install in the myenv.yml by using myenv.add_pip_package("fbprophet"). Then, the step of building the image failed.
                                    The error message did not have any information helpful for debugging. Although there might be somewhere with more detailed logs, but I had no clue where to find the logs. I had to google around for any possible reason.

                                    After being blocked by this error…

                                    1 vote
                                    Vote
                                    Sign in
                                    Check!
                                    (thinking…)
                                    Reset
                                    or sign in with
                                    • facebook
                                    • google
                                      Password icon
                                      Signed in as (Sign out)
                                      You have left! (?) (thinking…)
                                      0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                    • Trail between experiment and models

                                      We would like a way of tracing a model back to the experiment run that created it.

                                      1 vote
                                      Vote
                                      Sign in
                                      Check!
                                      (thinking…)
                                      Reset
                                      or sign in with
                                      • facebook
                                      • google
                                        Password icon
                                        Signed in as (Sign out)
                                        You have left! (?) (thinking…)
                                        0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                      • PYTHON into Azure

                                        Would be great to see Microsoft investing in Python, in order to embedded it into the Azure Machine Learning.

                                        1 vote
                                        Vote
                                        Sign in
                                        Check!
                                        (thinking…)
                                        Reset
                                        or sign in with
                                        • facebook
                                        • google
                                          Password icon
                                          Signed in as (Sign out)
                                          You have left! (?) (thinking…)
                                          0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                        • Pivot and unpivot needed come on this is common datascience component

                                          Pivot and unpivot needed come on this is common datascience component

                                          1 vote
                                          Vote
                                          Sign in
                                          Check!
                                          (thinking…)
                                          Reset
                                          or sign in with
                                          • facebook
                                          • google
                                            Password icon
                                            Signed in as (Sign out)
                                            You have left! (?) (thinking…)
                                            0 comments  ·  Flag idea as inappropriate…  ·  Admin →
                                          ← Previous 1 3 4 5 20 21
                                          • Don't see your idea?

                                          Machine Learning

                                          Feedback and Knowledge Base