SQL for BI and Data Science

During this training, you will learn how to use SQL for analysing data. It may come as a surprise that so-called "in-database" analytics is used much less than it could be: most often, the (production) data is first copied to an analytics platform or BI tool to do the "data science" there. For large data volumes, or for data "in the cloud", this is not the best choice since precious time is lost by copying the data. This course will teach you how you can perform data analytics in the database itself. The learned techniques will of course also be applicable on data which has been copied, to an environment that supports SQL, e.g. to a data warehouse (DW), or to a Big Data environment like Hadoop (with Hive) or Spark.

After a short introduction on data warehouses, data lakes, business intelligence (BI) and Data Science topics, we dive into a thorough discussion on the syntactic possibilities offered by SQL. More specifically, statistical, analytical and OLAP functions will be treated, as well as the use of SQL for ETL, for data exploration, and (if applicable) for MapReduce. The most important ones are individually discussed and demonstrated; realistic labs should allow the participant to get an idea of the applicability of these SQL functions and options.

The SQL syntax covered in this course is platform independent. When relevant, platform specific syntax will be pointed out with respect to Oracle, Db2, SQL Server and Hive/Spark.

At the end of this training, participants will be able to:

• understand how SQL (ie. the SELECT instruction) can be used for reporting and explorative purposes in the context of DW, BI, Big Data and Data Science;
• write effective SQL from a DW / BI / Big Data perspective;
• write problem-based SQL statements, and evaluate possible alternative SQL constructs for their effectiveness