Decomposition Ensemble Forecasting with Choquet Fuzzy Integral and its Application to Forecast International Tourist Arrivals

Decomposition ensemble forecasting with nonlinear artificial intelligence methods has been widely used in tourism demand forecasting. Because of the high practicality of the Choquet fuzzy integral for performance evaluation in the field of multiple attribute decision making, the aim of this study was to investigate the Choquet fuzzy integral as an ensemble model in terms of its forecasting performance with decomposition ensemble learning. To develop more accurate decomposition ensemble models to forecast international tourist arrivals, we used neural networks and the Choquet fuzzy integral to forecast individual component series extracted from original tourist arrival series and combined individual forecasts, respectively. Inbound tourism demand in Taiwan was employed to appraise the accuracy of the proposed decomposition ensemble models. The results revealed that the proposed models performed significantly better than the benchmark models we considered. The proposed models can thus benefit policymaking for tourism practitioners.