Tourist arrival forecasting with game-theoretic social influence weighting and deep learning
The study developed a hybrid forecasting method for tourist arrivals in Hong Kong and Sanya. It combines game theory to identify influential social media leaders, Empirical Wavelet Transform to handle complex time-series data, and a Stacked BiLSTM network tuned by Bayesian Optimization. The results show this approach significantly outperforms existing models by effectively integrating big data and opinion leader insights.