MLP annconvergence test

Hi,

I’m a beginner in TMVA MLP training and have a question.
When I try to get output of MLP annconvergence test, Test and Train set convergence estimator have different normalization, while what I expecting(or learned) (below plot) was having same level.

It would be very appreciated if anyone could give me solution about this problem.

image

Thanks,
Iker

Hi @iker; I am sure @moneta can help you with that.

Cheers,
J.