Data Mining Resources for SQL Server Developers
Below are a few resources that I’ve discovered recently that show various ways of getting started with the Data Mining capabilities of SQL Server.
Enhancing Applications with SQL Server Data Mining by Peter Myers (TechEd Australia 2013)
In this excellent talk by Peter Myers, he demonstrates how easy it can be to build a data mining model then shows three different ways that they can be incorporated into an application:
- Embedding data mining visualizations into a WinForms application
- Using a data mining models as the source for a Reporting Services report, embedding a report into a WinForms application
- Using the mining model to validate the “likeliness” of data input in a WinForms application
In addition to giving a very practical demonstration of creating and using a data mining model he also gives an excellent overview of some of the fundamentals associated with data mining. In particular he gives an excellent overview of the various models that are available with SQL Server.
Advanced-and Easy!-BI with Excel and SQL Server Data Mining by Rafal Lukawiecki (SharePoint Conference 2014)
In this talk by Rafal Lukawiecki (b | t), he once again gives a general overview of Data Mining, explaining some of the concepts then shows how to get started using the (free) Data Mining plugin for Excel 2013. He then moves onto building models using SQL Server Data Tools.
Most of the talk is spent within Excel, which really does show it is capable of almost anything :)
Forecasting with the Microsoft Time Series Data Mining Algorithm by Peter Myers (TechEd 2014)
In this talk from TechEd, Peter Myers focuses on the time series data mining algorithm in depth. He shows how it can be used to do forecasting using historical data.
The Level 300 (Experienced) classification is definitely appropriate, still if you want an in depth example of using the Time Series algorithm this is the video for you.
Data Mining in SQL Server Analysis Services by Brian Knight (SQLPass 2013)
In this excellent entry level presentation Brian Knight (of Pragmatic Works) demonstrates how simple and easy it is to get started working with Data Mining. If you’re looking for an introduction to the art of the possible this is a great place to start.
Kaggle - Data sets
In the two talks above (and plenty others I’ve seen in the past) the data set used for the examples is the well known AdventureWorksDW, if you’re a bit bored of this and fancy something different take a look at Kaggle.com.
Kaggle, if you haven’t come across it already is “…the leading platform for predictive modeling competitions.”, various companies publish interesting data sets as part of competitions, which you could enter to win prizes and money, or use the data sets for experimentation (check the TOC).