Forecasting Short-Term Tourism Flow of Mountain Resorts Based on Adaptive PSO-SVR

According to small samples and nonlinear characteristic of mountain resorts,the article combines Support Vector Regression with Adaptive Particle Swarm Optimization,which uses superiority of SVR in small samples and nonlinear forecasting and APSO searching for SVR model parameters of optimization,to forecast short-term tourism flow.The daily data set of a 5A mountain resorts from 2008 to 2011 summer holidays in Mount Huangshan is applied as an example.The experimental results demonstrate that the APSO-SVR approach is an effective way to forecast short-term tourism flow with greater accuracy and few errors of all above models including those of PSO-SVR,GA-SVR and BPNN.