This article presents the system “Neural Connect 4”, a program that plays the game Connect Four. This system employs the multilayer perceptron architecture which is learning t...
The number of required hidden units is statistically estimated for feedforward neural networks that are constructed by adding hidden units one by one. The output error decreases w...
A new graph-based evolutionary algorithm named “Genetic Network Programming, GNP” has been proposed. GNP represents its solutions as directed graph structures, which can improv...
Real world processes with an “intensity” and “direction” component can be made complex by convenience of representation (vector fields, radar, sonar), and their processin...
Abstract. Learning algorithms relying on Gibbs sampling based stochastic approximations of the log-likelihood gradient have become a common way to train Restricted Boltzmann Machin...