encog-train.getError（）和network.calculateError（dataset）之间的差异

Question

I have used both functions when training a network and both give approximately same values for each epoch. They are not exactly the same. I used network.calculateError(trainingdata).

What is the difference between getError() function and calculateError() function? Does getError() gives MSE for 1 epoch during training?

Answer 1

train.getError Error is calculated BEFORE training is applied by the call to iteration. See the documentation here BasicTraining.getError()

Whereas network.calculateError() calculates the error of the network using a regression calculator. This is why they will differ slightly.