Research on E-Commerce Customer Segmentation Optimization Based on Hybrid Genetic-Simulated Annealing Algorithm

This study addresses the limitations of traditional clustering algorithms in processing high-dimensional sparse e-commerce data by proposing a Hybrid Genetic-Simulated Annealing algorithm for customer segmentation optimization. The algorithm integrates the global search capacity of genetic algorithms with the local optimization capability of simulated annealing, employs a TF-IDF weighting mechanism to address data sparsity, and incorporates a dynamic temperature decay function with the Metropolis criterion for enhanced performance. Experimental results on the dataset of one thousand e-commerce user behavior demonstrate significant improvements in clustering effectiveness compared to conventional methods.