A feedforward neural network based on multi-valued neurons is considered in the paper. It is shown that using a traditional feedforward architecture and a high functionality multi-...
- The model of a simple perceptron using phase-encoded inputs and complex-valued weights is presented. Multilayer two-input and three-input complex-valued neurons (CVNs) are implem...
Howard E. Michel, David Rancour, Sushanth Iringent...
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 multilayer neural network based on multivalued neurons (MLMVN) is a neural network with a traditional feedforward architecture. At the same time, this network has a number of spe...
Igor N. Aizenberg, Dmitriy Paliy, Jacek M. Zurada,...