In the United States, it is a holiday week, so the news is pretty limited from many of the big cloud providers. Luckily, Amazon has come through with a flurry of machine learning announcements. Amazon is holding their annual re:Invent Conference next week, so maybe these announcements are precursors to some bigger news next week. We will have to wait and see.
Announcements
- Amazon Athena and Aurora add support for ML in SQL Queries
You can now invoke Machine Learning models right from your SQL Queries. You can choose from the built-in models provided by Amazon SageMaker, or you can use your own models which have been trained and deployed with SageMaker. - Use Amazon Sagemaker to add ML predictions in Amazon QuickSight
Amazon QuickSight, the AWS BI tool, now has the capability to call Machine Learning models. - Amazon AutoML now provides details on all iterations
Previously, Amazon’s AutoML only provided details on the best model discovered. Now the AutoML will provide details on all model run iterations. - Amazon Comprehend launches real-time classification
Amazon Comprehend is a service which uses Natural Language Processing (NLP) to examine documents. Comprehend can now be used to classify documents in real-time. Document classification no longer needs to be performed in batch processes.
Videos
- Preparing and Curating your data for Machine Learning
A great video from Google. It covers 5 tips to make sure your data is ready for machine learning: labels, relevancy, bias, consistency, and data leakage. It is based upon this article: Preparing and curating your data for machine learning.
- AI Show – Deep Learning vs Machine Learning
Want to know the difference between deep learning and machine learning. This video does a good job of answering the question. It has a companion blog post: Deep Learning vs Machine Learning.
- Azure Machine Learning Compute Instance
What used to be called Notebook VMs, are now Compute Instances. This a is a great way to share the same compute infrastructure with other members of your team. Make sure to watch until the end of the video to see the compute sharing in action.
Thanks for reading.