Abstract:

This paper presents a comprehensive framework for automating bank transactions and invoice reconciliation through a multi-criteria optimization model coupled with neural networks. The proposed framework aims to minimize manual mapping efforts while improving the accuracy and efficiency of the reconciliation process. We delve into various optimization techniques, and their computational challenges, and propose a novel approach using utility functions for multi-criteria decision-making. Neural networks are utilized to enhance data processing and decision-making capabilities. The results demonstrate significant processing time and accuracy improvements, offering a scalable and efficient solution for financial operations.