problems with prediction by desicion tree

[anonymo...@sqlserverdatamining.com]

I have just created a Desicion tree for the Bikebuyer example from the turorial. When I run
my precition query on the prospect table all are predicted to be non-byers (0) with the same probability (0.50580180180180179).
This did not happen in the Beta version.
What is the problem now? (I have installed the new release)
Her is my query

SELECT
t.[ProspectAlternateKey],
[v Target Mail].[Bike Buyer],
PredictProbability([v Target Mail].[Bike Buyer])
From
[v Target Mail]
PREDICTION JOIN
OPENQUERY([Adventure Works DW],
'SELECT
[ProspectAlternateKey],
[LastName],
[Gender],
[YearlyIncome],
[TotalChildren],
[NumberChildrenAtHome],
[HouseOwnerFlag],
[NumberCarsOwned]
FROM
[dbo].[ProspectiveBuyer]
') AS t
ON
[v Target Mail].[Last Name] = t.[LastName] AND
[v Target Mail].[Gender] = t.[Gender] AND
[v Target Mail].[Yearly Income] = t.[YearlyIncome] AND
[v Target Mail].[Total Children] = t.[TotalChildren] AND
[v Target Mail].[Number Children At Home] = t.[NumberChildrenAtHome] AND
[v Target Mail].[House Owner Flag] = t.[HouseOwnerFlag] AND
[v Target Mail].[Number Cars Owned] = t.[NumberCarsOwned]

[jami...@online.microsoft.com]

A few things to check are to do explore data in the DSV to see that the data the model is trained with is correct, and to check the tree to see if the tree actually splits.

[anonymo...@sqlserverdatamining.com]

I have the exact same problem. I am running the trial enterprise edition of SQL Server 2005 on Windows Server 2003. I distinctly remember this tutorial working on the beta editions.

The Decision Tree algorithm in general seems to have problems. Several times, processing a DT model has caused SQL Server to crash.

~Eric

[jami...@online.microsoft.com]

Thanks - we haven't seen this happen, but will look into it.
Any additional information you can provide would be helpful, especially regarding crashes. (In fact, is it the tools or the server that crashed?)

Thanks
-Jamie

[Jamie MacLennan (MS)]

We looked into this a bit and found that the reason you see this may be that
the prospective buyers table doesn't have an Age column and Age is the first
split in the tree. Therefore all cases get stuck in the root node and
return those probabilities. One workaround is to remove the Age column from
the model.

We'll investigate as to why the prospective buyers table is out of sync with
the tutorial.

--

-Jamie MacLennan
SQL Server Data Mining
This posting is provided "AS IS" with no warranties, and confers no rights.
<anonymo...@sqlserverdatamining.com> wrote in message
news:8d597fd8-6280-486e...@msnews.microsoft.com...