Microsoft Details Azure AI and Machine Learning Improvements

A week ahead of its Build developer conference in Seattle, Microsoft is detailing progress and improvements to its Azure platform. The company has detailed steps in Azure with artificial intelligence and machine learning.

Attachment 24696

On the AI side, Microsoft’s Azure Cognitive Services will see a new category called “Decision,” which includes a content moderator and anomaly detector, to make decisions. A new service entitled Personalizer will make recommendations to users. AI will also be a part of Azure Search.

On the intelligent edge (a term for computing locally, rather than in the cloud), the company announced “IoT Plug and Play,” a language to let developers more easily and instantly connect internet of things devices to the cloud. There’s also Azure SQL Database Edge, a SQL engine without the computing requirements of the cloud, for local machines that can still use machine learning and MIcrosoft’s graph. The company said this will make it easier for developers to work with cloud applications locally, without having to use different tools or programming languages than the cloud.

In addition, Azure’s Machine Learning will gain some new features that should make models easier to build and train. These include automation, a visual, drag-and-drop machine learning interface for creating models without any code, and improved support for Nvidia’s TensorRT and Intel nGraph for hardware-accelerated machine learning models.