Hybrid Approach for Forecasting Tourist Arrivals

For the tourism industry, accurate forecasts of travel needs are essential to meeting relevant needs, providing pertinent information to the government, and enabling stakeholders to adjust plans and policies. This study devised an approach that combines feature selection and support vector regression with particle swarm optimization (FS-PSOSVR) to forecast tourists to Singapore. The monthly tourist arrivals to Singapore from January 1978 to December 2017 were utilized as a test dataset. The results showed that the error obtained through FS-PSOSVR was smaller than that through other methods, revealing that FS-PSOSVR is an effective method for predicting tourism demand.