From data preparation to a Machine Learning model leveraging R Server on Spark: a real-case

Luxottica represents a customer that decided to partner with Microsoft for its Advanced Analytics solution and bought Cortana Intelligent Suite. Since they believe in open source, they wanted to leverage Spark, SparkR and R.

The project was mainly divided into:

- Data preparation (i.e. joining and aggregating) we use the data manipulation functions contained in the SparkR package.

- Modelling we use Microsoft R Server's ScaleR package to train models using the Spark Engine.

- Operationalizing we publish the model to Azure ML as a web service.

The demo will be done by using different datasets but will show the same principles and technologies applied for the real customer case.