Tourism demand forecasting by improved SVR model

The inboard tourism demand forecasting is very important to the development of tourism industry. In this paper, the SVR model is adopted to forecast monthly inbound tourism demand of China. And the elitist Non-dominated Sorting Genetic Algorithm (NSGAII) is used to parameter optimization. The NSGAII algorithm can reduce complexity of the algorithm, keeps the diversity of population and increasing the forecasting accuracy. At last, the proposed NSGAII-SVR model is used to forecast monthly inbound tourism demand of China from July 2011 to December 2011. And the experimental results show that the NSGAII-SVR has the best performance on forecasting compared with other models. © 2015 SERSC.