I have been able to use ScaledConjugateGradient when training an ANN but when attempting to use CrossValidationKFold, it no longer works. I have been able to get CrossValidationKFold working with ResilientPropagation and I noticed that ResilientPropagation and CrossValidationKFold work by copying a folded flat network to the referenced flat network used by the trainer. ResilientPropagation uses the iteration method defined in Propagation, but ScaledConjugateGradient overrides that method so that may be where the issue is...
I have been able to use ScaledConjugateGradient when training an ANN but when attempting to use CrossValidationKFold, it no longer works. I have been able to get CrossValidationKFold working with ResilientPropagation and I noticed that ResilientPropagation and CrossValidationKFold work by copying a folded flat network to the referenced flat network used by the trainer. ResilientPropagation uses the iteration method defined in Propagation, but ScaledConjugateGradient overrides that method so that may be where the issue is...