NARX closed-loop training

Hi Greg,

I can not find in the MATLAB documentation whether backpropagation-through-time or real time recurrent learning algorithms are used in the training of closed-loop NARX networks. I presume that the open-loop (series-parallel) configuration uses standard backpropagation, therefore, I would correspondinly believe they used backpropagation-through-time to update the weights and biases in the parallel configuration.

Kind Regards Juan.