Design and effectiveness verification of cross-border e-commerce tariff avoidance algorithm based on federal reinforcement learning and digital twin: an empirical study on China-US tariff game

作者：Zihao Deng · 年份：2025 · DOI：10.1117/12.3092649 · 被引用次数：1 · 研究领域：E-commerce and Technology Innovations、Advanced Technologies in Various Fields、Applied Advanced Technologies

Against the backdrop of the digital economy driving the reconstruction of the global trade landscape, this study constructs a three-dimensional analytical framework of "core population-supporting population-external environment" based on business ecosystem theory, coupling coordination theory, and digital trade innovation theory, systematically analyzing the empowerment mechanism of cross-border e-commerce on new-quality productivity. By integrating panel data from 30 provinces in China from 2015 to 2022 and applying the entropy method, Spatial Durbin Model (SDM), and improved three-system coupling coordination degree model (CDI index), this study reveals the spatiotemporal evolution laws and institutional barriers of cross-border e-commerce ecosystems.