A fast learning algorithm for neural network applications
A.S. Pandya, Raisa R. Szabo · 2002
Describes the use of the ALOPEX algorithm for solving nonlinear learning tasks by multilayer feedforward networks. ALOPEX is a stochastic parallel process. They demonstrate the use of ALOPEX for modifying the weights in a multilayer perception using a measure of global performance of the network. It estimates the weight changes by using only a scalar cost function which is a measure of global performance. The results of computer simulations of applying ALOPEX to nonrecurrent networks which include any feedforward architecture, in addition to multilayer perceptrons, are presented.>