Cloud Data Science News – Beta #5

Huge week of machine learning news from Amazon. This week Amazon hosted the large AWS re:Invent Conference. And there are…tons….of machine learning announcements from that event.

Announcements

• Amazon SageMaker Studio
  A browser-based Integrated Development Environment (IDE) for machine learning. This is going to be very helpful.
• Amazon SageMaker Experiments
  Allows for organizing, tracking, and comparing of machine learning models. It tracks things like: parameters, hyperparameters, and results.
• Amazon SageMaker Debugger
  Just what you would think, plus, it helps you prioritize the features which are having the greatest impact on the model performance.
• Amazon SageMaker Model Monitor
  Monitoring production models for data drift.
• Amazon SageMaker AutoPilot
  This is AutoML with full control and visibility.
• Amazon Braket
  An environment for quantum computing.
• Netflix and AWS open source Metaflow
  Making it easy to build and manage real-life data science projects. This will be an exciting project to watch.
  

Videos

• AWS re:Invent Machine Learning Announcements
  AWS CEO details all of the Machine Learning announcements during his keynote. Announcements start around the 1:33:40 mark.

• Democratize AI with Azure Machine Learning designer
  How do you select the correct machine learning algorithms? What is the new Azure Machine Learning Designer. Those questions are answered in this video.

• Azure Arc Announcement from Ignite 2019
  Azure Arc allows anyone to run Azure Data services on any hardware. That can be on any cloud, on-prem devices, or even IoT devices. It allows for a truly hybrid computing infrastructure. This announcement came about a month ago but this video clip just went live this week.

Learning

• What is Microsoft Learn?
  An amazing free learning resource. Hundreds of free courses.

• AWS Announces Machine Learning Embark
  The program helps organizations train existing employees on machine learning. It is a 3-month program. This is not free and requires acceptance.
• The Amazon Builders’ Library
  This is an organized grouping of articles about how Amazon builds and operates software.