A Time Series Forecasting Model Using a Hybrid Fuzzy Neural Network and SARIMA
Deshan Sun · Computer Technology and Development · 2008
The fuzzy neural network and season autoregressive integrated moving average(SARIMA) model separately have the extraordinary forecasting ability to the non-linear and linear time series.But most time series are not stable,simply linear or non-linear series in practical application.In order to impove precision forecasting,proposes a time series forecasting model using a hybrid fuzzy neural network(FNN) based on Takagi-Sugeno(T-S) and SARIMA.According to the experiment data of Sydney Airport traffic revenue passengers,three kinds of hybrid models are used and compared among four other models,i.e.,the FNN model,the Support Vector Machines model,SARIMA and the BP neural network model.The experiment result indicates that the proposed model is effective in the parameter choice and precision forecasting.