Deep Learning Coupling Analysis Model of Customer Topology Characteristics and Organizational Energy Efficiency

This study constructs a coupling analysis model of customer topology characteristics and organizational energy efficiency based on deep learning technology, revealing the dynamic impact mechanism of customer network structure on enterprise resource utilization efficiency. The topological features such as customer network centrality and modularity are extracted through graph neural network (GNN), and the temporal convolutional network (TCN) is combined to capture the temporal dependency. The model (MSE=0.012, R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>=0.823) is significantly better than traditional methods in prediction accuracy and generalization ability. The empirical results show that customer network centrality improves energy efficiency through resource integration effect (manufacturing <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\beta=0.42^{\ast\ast}$</tex>), while excessive modularization inhibits energy efficiency (service industry <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{\beta}=- 0.33\ast\ast$</tex>). Heterogeneity analysis shows that enterprise scale and ownership structure significantly regulate the above effects. Large enterprises benefit from network synergy advantages (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\beta=0.38 \ast\ast$</tex>), and state-owned enterprises weaken the negative impact of modularization due to policy support (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\beta=-0.12$</tex>). The study provides theoretical support and practical basis for enterprises to optimize customer structure and policy makers to design industry differentiation strategies.