Hybrid neural networks for time series forecasting

Abstract

The paper presents research in the field for hybrid neural networks for time series forecasting. A detailed review of the latest researches in this field is described. The paper includes detailed review of studies what compared the performance of multiple regression methods and neural networks. It is also consider a hybrid method of time series prediction based on ANFIS. In addition, the results of time series forecasting based on ANFIS model and com-pared with results of forecasting based on multiple regression are shown.