A Rickety Stairway to SQL Server Data Mining, Part 14.6: Custom Mining Functions

In the last installment of this amateur series of mistutorials on SQL Server Data Mining (SSDM), I explained how to implement the Predict method, which controls how the various out-of-the-box prediction functions included with SSDM are populated in custom algorithms. This limits users to returning the same standard columns returned by calls to prediction functions against the nine algorithms that ship with SSDM. The PredictHistogram function, for example, will return a single prediction column, plus columns labeled $Support, $Probability, $AdjustedProbability, $Variance and $StDev, just as it does with the out-of-the-box algorithms. Although we can control the output of the standard prediction functions to return any values we want, it may be more suitable in some cases to specify our own output format with more flexibility. This is where the ability to define your own custom mining functions comes in handy.