Neuro-fuzzy vs Neural Network Forecasting
Hubert Fernando, Lindsay W. Turner · CAUTHE 2006: To the City and Beyond · 2006
This paper consists of the application of a new hybrid combination of fuzzy logic and neural networks to forecast tourist arrivals to Japan. The Adaptive Neuro-Fuzzy Inference System (ANFIS), which is new to tourism studies, is used to make univariate and multivariate forecasts. These forecasts are compared with those of univariate and multivariate Multi-Layer Perceptron (MLP) neural network models. While MLP models show greater forecasting accuracy the ANFIS models have demonstrated sufficient credibility to justify further research in its application to tourism forecasting. This research is important to tourism forecasting because it explores a new approach to forecasting tourism.