Telecommunications data forecasting based on a dynamic neuro-fuzzy network

Abstract

In this work a dynamic neuro-fuzzy network (DyNF-Net) is proposed, which is applied on the outgoing telephone traffic of a large organization. It is a modified Takagi-Sugeno-Kang fuzzy neural network, where the consequent parts of the fuzzy rules are neural networks with internal recurrence, thus introducing dynamics to the overall system. Real world telecommunications data are used in order to compare the DyNF-Net to well-established forecasting models. The comparison highlights the particular characteristics of the proposed neuro-fuzzy network. © 2011 Springer-Verlag.