Noob question: why should we normalize test data with mean and std from training data? • /r/MachineLearning

Open in new window