Container Classification: A Hybrid AHP-CNN Approach for Efficient Logistics Management

Abstract

This research presents a groundbreaking approach that integrates artificial intelligence (AI) and big data for container classification, utilizing the Analytic Hierarchy Process (AHP) and Convolutional Neural Network (CNN). The study aims to address the challenges associated with prioritizing containers based on weight, destination, special requirements, financial considerations, and additional criteria.

The multi-criteria AHP method is employed to determine the relative importance of each criterion, providing weighted inputs for the subsequent CNN classification. The hybrid AHP-CNN model is strategically designed to optimize container classification, minimizing reshuffling movements within container yards, and facilitating efficient prioritization.

Through a comprehensive simulation, the effectiveness and adaptability of the proposed model are showcased. The study includes a sensitivity analysis, evaluating the accuracy of the model across various weight scenarios. The results demonstrate the robustness of the hybrid model, achieving a high level of accuracy in container classification. Notably, in three distinct scenarios, the model exhibited accuracy rates of 89.00%, 88.84%, and 91.05%, respectively.