Normalisation of Test set based on Mu and Sigma of Train set

[elijah981]

Hi Levent,

Excellent tutorial. It was very helpful.
I have one observation though, please correct me if I am wrong.

The normalization of the test set should be done with the mean and std of the train set, right?
Otherwise, if one is just doing prediction for one value, then the result wouldn't be representative.

Then, what it means is, one must store the mu and std of the train when it is normalized.
Let me know, what you think about it.

Thanks,
Elijah