I first divide the dataset into training (75%) and test (25%). Then fit a logistic regression model on training data set. When fitting the model, I did some modification on independent variable, such as normalization, transformation stuff. Now I have the model and ready to test it using test data set. But do I need to do the same transformation to the independent variable in the test data set?

  • 5
    $\begingroup$ Your fitted model links its input with your observations. If you transformed the input before fitting the model, it will not make sense for untransformed inputs. So: yes, you do need to apply the same transformation to the test set inputs. $\endgroup$ Jan 11, 2013 at 22:27


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.