With the rapid growth of global tourism, Xiamen has drawn attention for its sustainable tourism development.This study applies GIS analysis, an improved CNN model, grey correlation analysis, and a Swin transformer to analyse the spatial distribution and influencing factors of Xiamen's tourism industry.The enhanced CNN performs well in spatial analysis, accurately identifying tourism resource distribution.Results show the improved CNN achieves an accuracy of 0.953, recall of 0.947, F1-score of 0.950, MSE of 0.039, and MAE of 0.201.The values are all superior to those of the multilayer perceptron (MLP) and the long short-term memory network (LSTM).The Swin transformer also excels in predicting employment impact and resource efficiency, with accuracies of 0.882 (energy consumption), 0.856 (resource recovery), 0.874 (water use efficiency), and 0.863 (waste management).Its performance is also superior to that of the vision transformer (ViT) and data-efficient image transformers (DeiT) models.The findings indicate the improved CNN effectively captures spatial distribution patterns, while the Swin transformer reliably predicts employment and resource utilisation outcomes.This research provides a valuable basis for policymaking and sustainable development of Xiamen.
Qinyou Li (Thu,) studied this question.