78 pages matching values in this book
Results 1-3 of 78
What people are saying - Write a review
We haven't found any reviews in the usual places.
GUY A DUMONT
Augmented MultiLayer Perceptron
6 other sections not shown
activation adaptive applications approach approximation architecture ART1 Artificial Neural Networks associative memory back-propagation Boltzmann machine cells circuit classification components computation connection weights Connectionist convergence corresponding defined dynamics equation error error function example feature feedforward Figure function given global gradient descent hidden layer hidden units Hopfield IEEE implementation initial input pattern inverse Inverse Kinematics iteration JAPAN learning algorithm learning rate linear mapping matrix method minimize multilayer Neocognitron neurocomputer neurons nonlinear number of hidden object obtained optimal output layer output units paper parallel Parallel Distributed Processing parameters pattern recognition perceptron performance pRAM prediction presented problem Proc processor propagation proposed represent representation robot Rumelhart sample self-organizing sequence shown sigmoid sigmoid function signal simulated annealing Singapore solution speech recognition stable step structure subnetwork supervised learning synaptic technique Theorem threshold topology training patterns training set updating values VLSI