Data Mining - Professional SQL Server Analysis Services 2005 with MDX - page 754

[v Target Mail].[English Education] =
t.[EnglishEducation] AND
[v Target Mail].[English Occupation] =
t.[EnglishOccupation] AND
[v Target Mail].[House Owner Flag] =
t.[HouseOwnerFlag] AND
[v Target Mail].[Number Cars Owned] =
t.[NumberCarsOwned] AND
[v Target Mail].[Commute Distance] =
t.[CommuteDistance] AND
[v Target Mail].[Region] = t.[Region] AND
[v Target Mail].[Age] = t.[Age] AND
[v Target Mail].[Bike Buyer] = t.[BikeBuyer]
The preceding prediction query is one of the variations of the DMX SELECT
query that has the following syntax:
SELECT [FLATTENED] [TOP <n>] <select expression list>
FROM <model> | <sub select> [NATURAL]
PREDICTION JOIN <source data query>
[ON <join mapping list>]
[WHERE <condition expression>]
[ORDER BY <expression> [DESC|ASC]]
The input data for prediction is specified after the keywords PREDICTION
JOIN. The <select expression list> contains the columns to be retrieved as
part of the results and includes columns from the input/case table and the
predicted columns which are specified after the SELECT keyword. The min-
ing model used for prediction is specified after the FROM keyword. The map-
ping of columns from input data set to the mining model attributes is specified
in the ON clause as seen in the preceding prediction query. The prediction
query retrieves four columns from the input table along with the predicted
column for each input row. Similar to executing MDX queries from SQL Serv-
er Management Studio, you can execute the preceding DMX query. You have
only learned a simple DMX query in this example. Analysis Services 2005
tools help you to build the DMX query graphically, but if you are the kind of
person who wants to write your DMX query this will be a good start. You can
learn more about DMX and writing prediction query from Analysis Services
2005 documentation.