Machine Learning on Azure

In a previous post we briefly talked about the “Data” offering from Microsoft, now called Cortana Intelligence Suite. In this article we touch on the Machine Learning part of their offering. A good general introduction can be found here.

There are some great examples on Machine Learning on Microsoft Azure, now called Cortana Analytics Gallery, recently renamed to Cortana Intelligence Gallery.


On the site you will find the following categories:

Solution Templates

These are end-to-end solutions for particular industry-dependent solutions. At the time of writing there weren’t many available yet, but we are sure this will grow. You can filter and search by different categories, such as Algorithms and Language.


Here you can find predictive analytic experiments that you can use to build your own models and that can solve interesting problems or that can be used to demonstrate machine learning techniques. These are ideal candidates to start from and use as practice labs.

Here you can find an overview.

Machine Learning APIs

Next to the experiments there are a lot of APIs that can easily be accessed. We tested one of these called Sentiment Analysis, you can read our findings here.

Browsing for solutions

You can use a dedicated faceted search to find a particular solution:


This is an interesting way of navigating through the different solutions available, for example drilling down to from Linear Regression to Neural networks gave following:


We noticed that some of the Language filters did not always gave correct results and it sometimes took a few seconds for the results to display.

But still, a very nice way of finding examples and getting around the ML world of Microsoft.