Moving advanced analytics to your SQL Server databases
Speaker: Enrico van de Laar
Duration: 1 hour
Traditionally advanced analytical solutions, like machine learning, require you to bring your relational data to the machine learning model. Your model would then perform a prediction and return the results. While the process described above is reliable, it involves moving your data between the database where it is stored and the location where your model resides. This also means an increase in the complexity of your analytical solutions. For instance, how do you trigger the scoring of new data as soon as it enters the database? Or, how can you design this process for real-time scoring?
With the release of SQL Server 2016 Microsoft integrated a solution to the questions above, in-database analytics, allowing you to bring the analytics to your data instead of the other way around. Through in-database analytics we can design, train and score models directly from SQL Server without moving data out and back into the database. This creates a huge advantage, especially when working with real-time predictions, but how do you implement in-database analytics in your environment?
In this session we are going to explore the various methods available inside SQL Server 2016 & 2017 to perform in-database analytics. From building and storing our models directly inside SQL Server, to performing real-time scoring on data as soon as the data is stored inside a table.
After this session you will be able to understand the advantages and disadvantages of the various in-database analytics methods and you will be ready to start building your first in-database models!