Tourism Demand Forecasting Considering Environmental Factors: A Case Study for Chengdu Research Base of Giant Panda Breeding

Abstract

Effectively prediction of the tourism demand is of great significance to rationally allocate resources, improve service quality, and maintain the sustainable development of scenic spots. Since tourism demand is affected by the factors of climate, holidays, and weekdays, it is a challenge to design an accurate forecasting model obtaining complex features in tourism demand data. To overcome these problems, we specially consider the influence of environmental factors and devise a forecasting model based on ensemble learning. The model first generates several sub-models, and each sub-model learns the features of time series by selecting informative sequences for reconstructing the forecasting input. A novel technique is devised to aggregate the outputs of these sub-models to make the forecasting more robust to the non-linear and seasonal features. Tourism demand data of Chengdu Research Base of Giant Panda Breeding in recent 5 years is used as a case to validate the effectiveness of our scheme. Experimental results show that the proposed scheme can accurately forecasting tourism demand, which can help Chengdu Research Base of Giant Panda Breeding to improve the quality of tourism management and achieve sustainable development. Therefore, the proposed scheme has good potential to be applied to accurately forecast time series with non-linear and seasonal features.