Research on Logistics Development Level Classification using Enhanced K-Means Clustering Algorithm based on AC-PSO Optimization

This paper introduces an optimized K-means clustering model based on AC-PSO, aiming to address challenges associated with the selection of initial cluster centers in traditional
$\mathrm{K}$
-means clustering. Leveraging the potent search capabilities of the Particle Swarm Optimization (PSO) algorithm, the AC-PSO algorithm overcomes issues present in traditional PSO, such as a lack of rigorous mathematical foundations and weakened search capabilities in later stages leading to susceptibility to local optima. The proposed algorithm incorporates adaptive update strategies and employs chaotic mapping-based inertia weight selection, providing a more robust optimization process and avoiding the pitfalls of local optima during the optimization journey. Comprehensive tests on benchmark datasets (Iris, Wine, and Breast Cancer) demonstrate the superior performance of the proposed model in terms of precision, fitness values, and convergence speed when compared to traditional K-means and K-means based on PSO. Furthermore, the application of the optimized model to the clustering of logistics development levels across 18 cities in Henan Province for the year 2020 validates its effectiveness in real-world scenarios. This optimized K-means clustering model offers a novel approach to address the challenge of initial cluster center selection, showcasing potential improvements in accuracy and efficiency across various clustering tasks.