https://106c4.wpc.azureedge.net/80106C4/Gallery-Prod/cdn/2015-02-24/prod20161101-microsoft-windowsazure-gallery/Microsoft.MachineLearningModelManagement.1.1.1-preview/Icons/Large.png

Machine Learning Model Management

Microsoft
Manage, deploy and unlock insights from your machine learning models serving in production.
https://106c4.wpc.azureedge.net/80106C4/Gallery-Prod/cdn/2015-02-24/prod20161101-microsoft-windowsazure-gallery/Microsoft.MachineLearningModelManagement.1.1.1-preview/Screenshots/Image01.png
https://106c4.wpc.azureedge.net/80106C4/Gallery-Prod/cdn/2015-02-24/prod20161101-microsoft-windowsazure-gallery/Microsoft.MachineLearningModelManagement.1.1.1-preview/Screenshots/Image01.png

Machine Learning Model Management

Microsoft
Manage, deploy and unlock insights from your machine learning models serving in production.

The preview Model Management capabilities of Azure Machine Learning allow you to create services with your machine learning models, and then monitor and manage the different versions of your models in production.

Leverage the Model Management capabilities along with Azure Machine Learning Experimentation capabilities to complete your data science workflow from data preparation and model creation through to model deployment and management in production.

The Model Management preview capabilities include:

  • Service Creation: Create or update a real-time or mini-batch service to serve your model recommendations to production traffic.
  • Image Creation: Create an image with the service that will serve your model. Deploy this on-premises or to another cloud. Re-use the image to replicate the service into multiple regions in Azure.
  • Model Data Collection: Turn on model data collection to collect inputs and predictions for your service in production. Use this data to monitor data drifts in production and retrain models.
  • Managing Versions: Monitor what services are serving which versions of a model at any time
  • Health and Diagnostics: Track the health of your service by turning on service diagnostics that flow into your Application Insights account created as part of Machine Learning Compute creation.
  • Auto-Scale: Turn on auto-scale to keep the target utilization of your Compute in check while addressing fluctuations in traffic.
https://106c4.wpc.azureedge.net/80106C4/Gallery-Prod/cdn/2015-02-24/prod20161101-microsoft-windowsazure-gallery/Microsoft.MachineLearningModelManagement.1.1.1-preview/Screenshots/Image01.png
https://106c4.wpc.azureedge.net/80106C4/Gallery-Prod/cdn/2015-02-24/prod20161101-microsoft-windowsazure-gallery/Microsoft.MachineLearningModelManagement.1.1.1-preview/Screenshots/Image01.png