How to useParallel in system identification?
Is there a way to use 'useParallel' on the nlarx function in the system identification, like with the train function in the deep learning toolbox? I'm using neuralnet as the nonlinearity. Here's the training section of my script: %% Training na = 3; nb = [3 1 1]; nk = [0 0 0]; net = feedforwardnet(10); net.trainParam.epochs = 5000; nonlinearity = neuralnet(net); sys = nlarx(zt, [na nb nk], nonlinearity);
John Michell answered .
2025-11-20
Currently there is no way to parallelize model estimation in System Identification Toolbox.