Re: mining process flow

The simplest way to to this is to build a a query in the BI Dev Studio tool
click the Save Query Result button.
The query should be something like
SELECT T.Key, Cluster() AS [Cluster] FROM YourModel PREDICTION JOIN
OPENQUERY(<training data reference>) AS T...

Another more scalable solution is to use the Data Mining Query task in
Integration Services to execute the same kind of query and save the results
in a database.

Your query should return at least the key column from the training data and
the cluster label.



If you are using SQL Server 2008 you can simplify the process and build the
LogReg models directly on top of the DMX query, without saving the results
in a separate table. Also, you could have all the LogReg models in the same
mining structure, using the DMX Filtering feature.
More details :
- for training a model only with cases in a specific cluster:
http://sqlserverdatamining.com/ssdm/Default.aspx?tabid=102&Id=24
- for DMX filtering: http://www.bogdancrivat.net/dm/archives/17




--
--
--
This posting is provided "AS IS" with no warranties, and confers no rights.
Please do not send email directly to this alias. It is for newsgroup
purposes only.

thanks,
bogdan

"Paulo" <paulo@xxxxxxxxxxxxx> wrote in message
news:96E9237D-E7B1-42BF-9879-EC99424C1CBC@xxxxxxxxxxxxxxxx
Hi guys

I need to save the result clusters in a table by code, how can I do that?

I'm creating a process flow, using sql data mining structures. This
process consist in run clusters and for each cluster, run a logistic
regression. The mining strucuture for that is already, but I need to find
a
way to save the cluster groups in a specific table.
I made a mistake with the name of this group but I know that you guys are
experts in SQL, and can help me to solve this trouble.

best
--
Paulo Carvalho
Statistics Master
PhD Markovs Chain
Univ S.Paulo


.

Loading