In July, Microsoft plans to unveil a preview of its Machine-Learning-as-a-Service platform, hosted on Windows Azure. The platform is expected to leverage Microsoft’s powerful Azure cloud platform to provide powerful predictive analytics services and off premises data crunching.
Microsoft calls this new platform Microsoft Azure Machine Learning. Interested developers and businesses can sign up at the site to receive upcoming news and announcements as the unveil date approaches.
Machine Learning is a form of simplistic (but powerful) artificial intelligence used for numerous data oriented analysis and is a mainstay of big data predictive analysis. Search engines, recommendation engines, GPS systems, and even expert systems such as Cortana make use of machine learning to provide better results to customers. Google has been in the news recently for its use of Machine Learning to make its data centers more power efficient as well.
Microsoft claims to be “democratizing machine learning” with Azure ML by putting it into the cloud and leveraging it as AI-as-a-service. By bringing together powerful algorithms produced for products such as Bing and Xbox Live and rolling it together in an easy-to-use cloud service, Microsoft is putting together a strong lineup of tools.
Azure ML comes with a design studio aimed at business analysts; an application programming interface (API) for developers to tap into; and a software development kit (SDK) to allow developers to build applications on top of Azure ML.
Azure ML early preview by example
Azure ML become available for preview next month, but some developers have an early preview and have been flexing it’s prediction capabilities already.
Microsoft’s TechNet article includes a few examples of how the early preview version of Azure ML is currently being used. For example, developers at MAX451 are using Azure ML to help a large retail client determine what customers are most likely to purchase next, this is based on vast amounts of data obtained from ecommerce and retail sales. And, OSISoft developers are using Azure ML to make Carnegie Mellon University’s campus more energy efficient by predicting faults and optimizing power.
No time frame has been set by Microsoft as for when Azure ML will come out of preview. For more information on the product visit the Technet blog post on Azure ML and Microsoft’s newly minted Machine Learning blog.