Find more posts tagged with
Sort by:
1 - 5 of
51
Hi!
You connected Group Models in a wrong way.
The first input of Group Models should be the pre output of Nominal to Numerical. The second will be the actual model (mod output of Linear Regression). Then you will apply the grouped model (the one consisting of NomToNum and the regression) using Apply Model. This makes sure that the same kinds of transformations are applied on both the training and the testing dataset.
Regards,
Balázs
You connected Group Models in a wrong way.
The first input of Group Models should be the pre output of Nominal to Numerical. The second will be the actual model (mod output of Linear Regression). Then you will apply the grouped model (the one consisting of NomToNum and the regression) using Apply Model. This makes sure that the same kinds of transformations are applied on both the training and the testing dataset.
Regards,
Balázs
Sort by:
1 - 1 of
11
You can use "Nominal to Numerical" to encode the categorical data to 0/1 numbers, by creating new attributes. Dummy coding is a good choice.
Remember that the linear regression doesn't work with nominal attributes, the model is waiting for numerical data.
Best,
Cesar