LOGISTICS TRANSPORTATION COST FUNCTION OPTIMIZATION

Abstract

Purpose of the study. Optimizing transportation costs is a fundamental challenge in logistics management that requires advanced techniques to handle the complexities of geographical distances, transportation modes, and operational constraints. This article investigates the use of RMSProp, RMSProp with gradient clipping, and Proximal Gradient Descent methods in the optimization of transportation cost functions within logistics networks. By incorporating quality functions associated with both origin and delivery points, the study seeks to achieve a balance between cost reduction and the enhancement of service quality. Methodology. We present a comparative analysis of the performance of these methods, focusing on their efficiency in terms of convergence rates and the quality of the solutions obtained. The research demonstrates that RMSProp and its variant with gradient clipping are particularly effective in navigating the solution space, offering fast convergence and high-quality solutions. Proximal Gradient Descent, on the other hand, shows promise in handling the discrete nature of logistics problems. Scientific novelty. This study underscores the critical role of transportation cost optimization amidst the growing demands of global trade and logistics. As global trade expands, the need to minimize transportation expenses while maintaining service quality becomes increasingly vital. Advanced optimization techniques like RMSProp and Proximal Gradient Descent can lead to significant cost savings and improved operational efficiency, benefiting businesses involved in international trade. Conclusions. In conclusion, this study highlights the effectiveness of RMSProp and Proximal Gradient Descent in optimizing transportation costs. It emphasizes the necessity of continuous innovation in logistics management to meet the evolving demands of global trade. Future research directions include the exploration of hybrid optimization techniques that combine the strengths of gradient-based methods, further enhancing the robustness and applicability of transportation cost optimization models in diverse logistical environments.