A supervised learning-based optimization for container pre-loading problem

This study proposes a novel supervised learning-based optimization algorithm to address the container pre-loading problem faced by manufacturing firms using third-party logistics (3PL) providers. The primary challenge of this problem arises from the significant variability in the weight of trucks managed by 3PL providers. To address this issue, our methodology incorporates supervised learning algorithms into the optimization process, leveraging truck weight predictions to efficiently minimize associated costs. Using real-world data from a leading beverage manufacturer, our algorithm demonstrates significant cost reductions and improvements in operational efficiency over other conventional benchmarks. Moreover, our research not only introduces a novel approach to the container pre-loading issue but also expands the potential for applying supervised learning-based optimization methods in diverse areas, offering valuable insights and practical benefits to the field.