Here are this week’s news and announcements related to Cloud Data Science. Plus, there are some links for Videos and Tutorials.
- Google Introduces Explainable AI
Many industries require a level of interpretability for their machine learning models. Black box solutions are not always ok. Google is launching Explainable AI which quantifies the impact of the various factors of the data as well as the existing limitations.
- Model Cards from Google
This is a brilliant idea. Google is beginning to make single page “cards” for common machine learning tasks. Each card contains a description, pros, cons, limations and examples for a specific machine learning task. Think about how foods put nutrional value and ingredients on the packaging. This is a similar concept, but for machine learning. There are currently examples for face detection and object detection.
- AWS Storage Day
On November 20, 2019, Amazon held AWS Storage Day. Many announcements came out regarding storage of all types at AWS. S3 got some enhanced features. Elastic File Storage (EFS) is available in 4 new regions: Bahrain, Stockholm, Sao Paulo, and Hong Kong. An many more announcement. Much of this is in anticipation of AWS re:Invent, coming in early December 2019.
- PyTorch on Azure with streamlined ML lifecycle
Microsoft Azure supports the latest version of PyTorch. Azure Machine Learning provides an environment to train, scale, deploy and monitor solutions using PyTorch.
- Azure now has data center regions in Norway
Microsoft Azure is now available from the new cloud datacenter regions in Norway. Microsoft continues to advance its global network with 54 regions.
- Breaking free of datacenter legacy thinking
This video covers some of the latest from Microsoft Research. Ant Rowstron walks you thru some of the amazing technologies which will make up datacenters in the future. Fascinating Stuff!
- Azure Open Datasets
Azure will now host open datasets and it will allow you to easily use those datasets in Azure Machine Learning.
- What’s new in Azure Cognitive Services
Azure Cognitive Services bring the power of AI to all developers. All a developer needs to do is call an API. They allows devs to have the “ability to see, hear, speak, search, understand, and accelerate decision-making.”
Tutorials and Learning
- Azure AI and Machine Learning talk series
This is a series of videos, targeted at developers, for learning how to use AI and ML with Azure.
- AWS Snowball Edge – Getting Started
AWS Snowball Edge is a large-scale data migration and edge computing device. This is the first in a three-part online training series with Snowball Edge.