Updated: March 2, 2016
Predicts future sequence values for a specified set of sequence data.
PredictSequence(<table column reference>) PredictSequence(<table column reference, n>) PredictSequence(<table column reference, n-start, n-end>)
A <table expression>.
If the n parameter is specified, it returns the following values:
If n is greater than zero, the most likely sequence values in the next n steps.
If both n-start and n-end are specified, the sequence values from n-start to n-end.
The following example returns a sequence of the five products that are most likely to be purchased by a customer in the Adventure Works DW Multidimensional 2012 database based on the Sequence Clustering mining model.
SELECT PredictSequence([Sequence Clustering].[v Assoc Seq Line Items],5) From [Sequence Clustering]