Getting Beyond the Basics with Azure Machine Learning

ABSTRACT

Azure Machine Learning offers a low-code approach to performing machine learning tasks, but what happens when you need to do something a little more complex? In this talk, we will go beyond the Azure Machine Learning designer and show how to use Azure Machine Learning in a code-first environment. We will understand how the Pipeline metaphor works within code and use that to generate and deploy models, using the example of a code-first model performing batch inference. We will also review how Azure Machine Learning handles model registration and versioning. This talk assumes some basic familiarity with Azure Machine Learning and the Python language.

ADDITIONAL MEDIA

No recordings or additional media are available for this talk.

SLIDES

Click here to access the slides for this presentation.

The slides are licensed under Creative Commons Attribution-ShareAlike.

DEMO CODE

Click here to access demo code for this presentation.

The source code is licensed under the terms offered by the GPL.

LINKS & FURTHER INFO

Helpful Resources

• Train a basic Azure ML model with Python
• The Azure ML product team has a video showing how to build models via notebooks.
• What are Azure ML pipelines?
• Create and run ML pipelines with the AML SDK
• Batch scoring with Azure Machine Learning pipelines
• An example notebook which shows off ML pipelines
• Curated environments allow for faster startup due to having pre-installed packages.
• We didn't cover AutoML very much, but you can definitely use it in Azure ML pipelines.
• Making your notebook-style code production-ready
• The Microsoft MLOps repo
• The reference architecture used in the above MLOps demo
• Getting started with MLOpsPython
• Google's whitepaper on MLOps
• Microsoft's whitepaper on MLOps
• Writing data to Azure SQL DB from Azure ML
• MLflow integrations with Azure ML