The Agenda

Machine Learning is the process of extracting predictive or categorical relationships from large quantities of multiple sources of data. Most often,the Data Scientist uses a series of tools and processes to gain meaning from data – and coordination between other professionals is a manual affair. Microsoft’s Azure Machine Learning (Azure ML) brings together a completely on-line collaborative experimentation environment to build solutions, and a simple way to publish the results so that other code and processes can use them.Buck Woody from the Microsoft Machine Learning and Data Science team (MLADS) will show you not only how to use the Azure ML tool to create solutions, but also explain the entire Cortana Analytics Suite and where AzureML fits in. You’ll also learn the process for creating the solution. We’ll take a solution from source data to published output. At the end of this workshop, you’ll be able to:

Understand the Cortana Analytics Suite and when and where to use Azure ML 
Apply the Cortana Analytics Process (CAP) to a given solution
Understand and use the Azure ML Studio environment to create collaborative experiments and publish solutions
Get input data from on-premises, online, and cloud-based sources
Clean, transform, normalized and quantize your data
Build, score and evaluate a Predictive Model
Build and evaluate a categorical model
Publish and stage the predictive model as an Azure-based Service  

Requirements:

You’ll need a laptop with connectivity to the Internet
Optionally: Microsoft Excel (if you wish to consume the model in Excel)