Multilayer Feedforward Neural Network Based on Multi-Valued Neurons (MLMVN)

Abstract

In this Chapter, we consider one of the most interesting applications of MVN - its use as a basic neuron in a multilayer neural network based on multi-valued neurons (MLMVN). In Section 4.1, we consider basic ideas of the derivative-free backpropagation learning algorithm for MLMVN. In Section 4.2, we derive the error backpropagation rule for MLMVN with a single hidden layer and a single output neuron and then for the most general case of arbitrary number of layers and neurons in the layers. In Section 4.3, the Convergence Theorem for the MLMVN learning algorithm is proven. In Section 4.4, we consider in detail how MLMVN learns when solving a classification problem. We also consider in the same section how MLMVN solves a problem of the Mackey-Glass time series prediction. Concluding remarks are given in Section 4.5.