Evolving GMDH-neuro-fuzzy system with small number of tuning parameters
Yevgeniy V. Bodyanskiy, Олена Бойко, Yuriy P. Zaychenko, Galib Hamidov · 2017
In the paper a new class of evolving fuzzy networks is suggested, namely the evolving GMDH-neuro-fuzzy systems (Group Method of Data Handling). Their main advantage is a small number of tuning parameters that simplifies the training algorithms and cut training time comparing with classical evolving fuzzy neural networks (EFNN). The architecture and training algorithms for this neural network are considered. The experimental investigations of the proposed GMDH-neuro-fuzzy systems were carried out in the problem of stock prices forecasting and comparisons with GMDH algorithm and the cascade neuro-fuzzy network were performed that enable their efficiency estimation.