N. V. Gridina, I. A. Evdokimov, V. I. Solodovnikov, “Construction of hybrid neural networks using fuzzy logic elements”, Artificial Intelligence and Decision Making, 2019, Issue 2,Pages <nobr>91

Evolutionary computation and soft computing

Construction of hybrid neural networks using fuzzy logic elements

N. V. Gridina^ab, I. A. Evdokimov^a, V. I. Solodovnikov^a

^a Center of Information Technologies in Design, Russian Academy of Sciences, Odintsovo, Moscow Regions, Russia
^b I. M. Sechenov First Moscow State Medical University, Moscow, Russia

Abstract: The issues of using fuzzy logic elements in the construction and training of hybrid neural networks for proceeding under conditions of uncertainty are considered. An embodiment of the fuzzy relations implementation and the fuzzy compositional output algorithm based on the neural network approach are presented. A neural network model of a fuzzy system is presented as a universal approximator, as well as the case while fuzziness appears at the learning stage.

Keywords: hybrid neural network; fuzzy logic; fuzzy compositional neural networks.

DOI: 10.14357/20718594190209