Fuzzy and chaotic neuro-network modeling
Algis Garliauskas · 2002
The origin of fuzzy signals lies in the complex biochemical and electrical processes of the synapse and dendrite membrane excitation and in the inhibition mechanism. The mathematical operations included in fuzzy neural network modeling are described. The scalar product between the inputs of layers and synaptic weights is replaced by a fuzzy logic multiplication. The sum of products changes into fuzzy logic sums, and operators such as supremum, maximum and minimum are presented for fuzzy description. Both fuzzy properties and a chaos phenomenon are analyzed based on experimental computations.