caa a22 4500 47580113X CHVBK 20180406123737.0 cr unu---uuuuu 170329e20000401xx s 000 0 eng 10.1023/A:1009682730770 doi (NATIONALLICENCE)springer-10.1023/A:1009682730770 Single-Iteration Training Algorithm for Multi-Layer Feed-Forward Neural Networks [Elektronische Daten] [J. Barhen, R. Cogswell, V. Protopopescu] A new methodology for neural learning is presented. Only a single iteration is needed to train a feed-forward network with near-optimal results. This is achieved by introducing a key modification to the conventional multi-layer architecture. A virtual input layer is implemented, which is connected to the nominal input layer by a special nonlinear transfer function, and to the first hidden layer by regular (linear) synapses. A sequence of alternating direction singular value decompositions is then used to determine precisely the inter-layer synaptic weights. This computational paradigm exploits the known separability of the linear (inter-layer propagation) and nonlinear (neuron activation) aspects of information transfer within a neural network. Examples show that the trained neural networks generalize well. Kluwer Academic Publishers, 2000 virtual input layer nationallicence neural network training nationallicence fast learning nationallicence SVD nationallicence Barhen J. Center for Engineering Science Advanced Research, Oak Ridge National Laboratory, P.O. Box 2008, Bldg. 6012, TN 37831-6364, Oak Ridge, USA aut Cogswell R. Department of Mathematics and Computer Science, Monmouth College, 61462, Monmouth, IL, USA aut Protopopescu V. Center for Engineering Science Advanced Research, Oak Ridge National Laboratory, P.O. Box 2008, Bldg. 6012, TN 37831-6364, Oak Ridge, USA aut Neural Processing Letters Kluwer Academic Publishers 11/2(2000-04-01), 113-129 1370-4621 11:2<113 2000 11 11063 https://doi.org/10.1023/A:1009682730770 text/html Onlinezugriff via DOI 1 research-article jats NATIONALLICENCE 856 40 https://doi.org/10.1023/A:1009682730770 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Barhen J. Center for Engineering Science Advanced Research, Oak Ridge National Laboratory, P.O. Box 2008, Bldg. 6012, TN 37831-6364, Oak Ridge, USA aut NATIONALLICENCE 700 1- Cogswell R. Department of Mathematics and Computer Science, Monmouth College, 61462, Monmouth, IL, USA aut NATIONALLICENCE 700 1- Protopopescu V. Center for Engineering Science Advanced Research, Oak Ridge National Laboratory, P.O. Box 2008, Bldg. 6012, TN 37831-6364, Oak Ridge, USA aut NATIONALLICENCE 773 0- Neural Processing Letters Kluwer Academic Publishers 11/2(2000-04-01), 113-129 1370-4621 11:2<113 2000 11 11063 Metadata rights reserved Springer special CC-BY-NC licence nationallicence BK010053 XK010053 XK010000 NATIONALLICENCE NATIONALLICENCE NL-springer