STANDARDIZATION

[s-tete]

Why are you using the mean and standard deviation of the training set to standardize the testing set in ln 61

[harimkang]

Why are you using the mean and standard deviation of the training set to standardize the testing set in ln 61

Hi @s-tete, thanks for creating issue,
This is because the same regularization process applied to the training set should be applied to the test, so that the network can make predictions in the same space. This assumes that the test and train sets are somewhat related.