Analysis of Provincial Export Performance in Türkiye: A Spectral Clustering Approach
创建于 更新于
摘要
This paper applies spectral clustering (with Eigen-Gap selection and Silhouette validation) to 2023 import, export and net-export data (raw and exchange-rate adjusted) for Türkiye's 81 provinces, finding an optimal 3-cluster partition (Low/Medium/High) and revealing pronounced regional disparities — e.g., İzmir has the highest net-export performance while İstanbul records the lowest net-export values; exchange-rate adjustments materially change some province rankings [page::0][page::11].
