Toward the training of feed-forward neural networks with the D-optimum input sequence | IEEE Journals & Magazine | IEEE Xplore