Combining Machine Learning and Metaheuristics Optimizing for 3PLs’ Daily Payment Schedules in Logistics

Abstract

In today's competitive business environment, logistics processes that encompass critical stages from procurement to customer delivery play a central role in supply chain management. Effective supplier management is crucial for gaining a competitive advantage, enhancing quality, and ensuring customer satisfaction. This study proposes an artificial intelligence-based model for generating daily payment schedules for third-party logistics providers (3PLs), which are a key component of logistics operations. The proposed model consists of two stages. In the first stage, 3PLs are scored based on their logistical capabilities and operational data using machine learning methods. In the second stage, daily payment schedules are automatically generated using a metaheuristic approach based on these scores and financial data from the payment system. The machine learning and metaheuristic methods used in the construction of the model were determined using logistical operational and financial data from Alışan Logistic for the period 2021-2023. The results showed that CatBoost Regression was the most successful method for scoring 3PLs, while the Genetic Algorithm was the most effective for generating payment schedules.