Re: Tree model parameter setting -- Force Regressor
- From: "Jamie MacLennan \(MS\)" <jamiemac@xxxxxxxxxxxxxxxxxxxx>
- Date: Fri, 18 Nov 2005 15:01:31 -0800
That's a confusing bit of UI. The Remove button is there for the case where
an algorithm has parameters that aren't exposed in the algorithm's metadata,
or when an algorithm can accept a parameter multiple times. This
functionality isn't used for the Microsoft algorithms, but can be used for
plug-ins.
If the value of a parameter is blank, it means the parameter is not set.
The issue you are seeing is that the REGRESSOR flag on the model column is
set. To clear this flag, select the Bike Buyer column under your Target
Mailing via Tree model on the Mining Models panel. The column properties
will show in the properties panel. The flag to clear is called Modeling
Flags
Let me know if there's any other issues.
Thanks
--
-Jamie MacLennan
SQL Server Data Mining
This posting is provided "AS IS" with no warranties, and confers no rights.
"ImJoe" <ImJoe@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:AFB5C528-C92E-43B2-9A7B-186FD6B87C08@xxxxxxxxxxxxxxxx
>
> This algo param has been checked to remove via the REMOVE button.
> However,
> the button action (of the REMOVAL) does not seem to have any impact (even
> another action of SAVE is followed). When going back to see the paras,
> this
> para (Force Regressor) is still there. Another strange thing about this
> para
> is, it's all blank for Value, Default and Range.
>
> Under this circumstance, a Predict Column with type of 'Discrete'
> generates
> the following err:
> Error 1 Error (Data mining): In the mining model, Target Mailing via
> Tree,
> the Bike Buyer column must be of a continuous type since the REGRESSOR
> modeling flag is set on it. 0 0
>
> Please advise a solution.
>
> TIA.
.
- Prev by Date: Re: Problem training model from script
- Next by Date: Re: Realtime Data Mining
- Previous by thread: Re: Problem training model from script
- Next by thread: Re: Tree model parameter setting -- Force Regressor
- Index(es):
Relevant Pages
|