High-End Hotel Location Evaluation and Prediction in Nanjing City: A Data-Driven Approach Using Multi-Source Spatial Data and Machine Learning

This study presents a data-driven approach to evaluating and predicting high-end hotel locations in Nanjing City using multi-source spatial data and machine learning. By analyzing the spatial patterns of high-end hotels in Nanjing and constructing a feature matrix from 27 spatial factors, the Random Forest model predicts suitable future hotel locations. The model's performance demonstrates high predictive accuracy (AUC = 0.989). Results indicate that high-end hotels are clustered in central districts, particularly Xuanwu and Gulou districts. Furthermore, the suitability maps generated by the model identified several hotspot areas where new high-end hotels are likely to thrive. The model’s predictions also highlighted potential locations in emerging districts where current hotel density is low but future demand is expected to grow. This integrative approach enhances strategic planning in the hospitality industry, which offers practical insights for investors and urban planners seeking to optimize hotel locations in rapidly growing cities. The research contributes to bridging the gap between spatial data analysis and predictive modeling in hotel site selection.