Difference between KNIME and Microsoft Server Data Tools / SQL Server 2016


This is my first post as I am trying to evaluate which software package to use for my PhD.  I am primarily interested in classification of data using machine learning within a 30Gb SQL database.

I cannot find any comparisions of KNIME with the data mining/machine learning functions in Microsoft Server Data Tools / SQL Server 2016.

Does anyone have any references, website or comment they can provide.

Because I am using an SQL database it would seem that using Microsofts tools would be the immediate place to start, how does KNIME differenciate itself?






Hi gjhodgson,

You can easily connect to your database using the SQL Server Connector node in KNIME.

To learn more about the classification and predictive modelling capabilities of KNIME, please have a look at the Node Guide: https://www.knime.org/nodeguide



Overall I'd say KNIME is for the data analysts, SSAS is for (Microsoft-centric) coders. KNIME has a tonne of native algorithms, much more than SSAS has natively, and apart from everyone's darling R is also integrates Weka and RapidMiner, so it has a lot more code-free options lined up.



Many thanks for your responses - I have also found that the Microsoft solutions (via either Azure Machine Learning or Microsoft Server data tools) is poor at dealing with mixed data - KNIME really is the answer if you need to easily do data pre-processing.  Also, all of the Microsoft options are very poor at reporting - whereas it's a piece of cake to produce a scatter plot with associated algorithm.

I completely agree with the comment that KNIME is for the data analysts, SSAS is for (Microsoft-centric) coders.

I hope that this also helps others when they are considering which solution to adopt.

1 Like