Forecasting coastal tourism demand is crucial for promoting high-quality development of the marine economy. With the deepening of internet technology and the widespread adoption of its applications, online search indices can profoundly reflect the actual needs of tourists during their decision-making process. Therefore, utilizing network index to forecast coastal tourism demand is feasible. Taking Gulangyu Island in Xiamen as a case study, this paper employs a MIDAS model. It constructs a baseline ADL model, univariate and multivariate MIDAS model using high-frequency Baidu Index data and low-frequency Gulangyu tourist growth rates. These models forecast tourism demand for Gulangyu. Subsequently, the fitted data from these models undergo further prediction via the XGBoost machine learning algorithm. Finally, the Root Mean Square Error (RMSE) of each model's prediction results is compared. The results indicate that sudden health events significantly impact the accuracy of predictive models.Both the univariate and multivariate MIDAS models outperform the baseline ADL model in predictive capability.

Kangning Li, Jianhua Xiao. Mixed-Frequency Model Forecasting of Coastal Tourism Demand Based on Network Index. Academic Journal of Business & Management (2025), Vol. 7, Issue 11: 54-60. https://doi.org/10.25236/AJBM.2025.071108.

[1] Ghysels, Eric, and R. Valkanov . "The MIDAS Touch: Mixed Data Sampling Regression Models." Cirano Working Papers 5.1(2004):512-517.

[2] Clements, M.P., and A.B.Galvão:“Macroeconomic Forecasting with Mixed-Frequency Data: Forecasting US OutputGrowth” [J],Journal of Business and Economic Statistics,2008, 26(4): 546—554.

[3] LIU Jinquan, LIU Han, YIN Chong: Application of Macroeconomic Mixed-Frequency Data Modeling in China-An Empirical Study Based on the MIDAS Model [J]. Economic Science 2010,(05):23-34.

[4] LIU Han, LIU Jinquan: Real-time forecasting and short-term prediction of China's macroeconomic aggregates--an empirical study based on MIDAS model[J]. Economic Research,2011,46(03):4-17.

[5] ZHENG Tingguo, SHANG Yuhuang: Mixed-frequency forecasting analysis of China's GDP based on financial indicators[J]. Financial Research,2013,(09),16-29.

[6] LIU Han, WANG Yonglian: Improvement of forecasting accuracy based on mixed-frequency forecasting model--Taking inbound tourism as an example[J]. Journal of Intelligence, 2016, 35(09): 75-79.

[7] HIRASHIMA A, ONES J, BONHAM C S, et al: Forecasting in a mixed up world: Nowcasting Hawaii tourism[J]. Annals of Tourism Research, 2017,(63):191-202.

[8] QIN Meng, LIU Han: Baidu index, mixed-frequency modeling and tourism demand in Sanya[J]. Journal of Tourism,2019,34(10):116-126.

[9] Ghysels, E., A.Sinko, and R.Valkanov. “MIDAS regressions: Further results and new directions”[J], Econometric Reviews, 2007,26 (1) , 53-90.