Options
Applying Normalization Parameters of Training Set to Test Set
Newbie
Hi,
before training a Neural Network the training set has to be normalized, e.g. with z-score.
Before validating the model, one would also normalize the test set by using the normalization parameters obtained from the normalization of the training set. This is done to prevent information leakage through normalization.
In SKlearn this can be done by fitting a scaler on the training set (scaler_A.fit(training_data)) and normalizing the test set using the same scaler (scaler_A.transform(test_data))
How can this be achieved with the normalization block in RapidMiner?
Thanks!
before training a Neural Network the training set has to be normalized, e.g. with z-score.
Before validating the model, one would also normalize the test set by using the normalization parameters obtained from the normalization of the training set. This is done to prevent information leakage through normalization.
In SKlearn this can be done by fitting a scaler on the training set (scaler_A.fit(training_data)) and normalizing the test set using the same scaler (scaler_A.transform(test_data))
How can this be achieved with the normalization block in RapidMiner?
Thanks!
Tagged:
0
Best Answer
-
Options
MartinLiebig
Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,507 
RM Data Scientist
Hi,thats why the Normalize has this pre port. Its a preprocessing model which you can apply on the test set like a normal model.Best,Martin- Sr. Director Data Solutions, Altair RapidMiner -
Dortmund, Germany1
