Cluster of Vulnerable Municipalities in Mexico to facilitate the Creation of Coordination Mechanisms in the Humanitarian Responses through Machine Learning Techniques

Abstract

Cluster of Vulnerable Municipalities in Mexico to facilitate the Creation of Coordination Mechanisms in the Humanitarian Responses through Machine Learning Techniques. Coordination among actors in the supply chain is essential for ensuring a well-organized, efficient, and effective response to humanitarian crises, ultimately leading to better outcomes for those affected by disasters. Effective coordination ensures that resources, information, and aid are distributed efficiently and promptly to those in need during humanitarian crises. It also helps optimize resource allocation, prevent duplication of efforts, and ensure that aid reaches the right places at the right time. It supports identifying and mitigating risks in the supply chain, such as delays, bottlenecks, or disruptions, which can impact aid delivery. Coordination fosters better communication among stakeholders, enabling them to share information, collaborate on solutions, and make informed decisions. Precise coordination mechanisms help establish accountability among actors, ensuring that responsibilities are defined, monitored, and fulfilled. This document proposes an efficient logistics system capable of providing aid in a rapid and coordinated manner through unsupervised learning and a medoid partitioning algorithm called PAM (Partitioning Around Medoids) that achieves a global optimum for grouping the vulnerable municipalities of Mexico using INECC data. In addition, the PAM algorithm was compared with the P-Median to check the efficiency of the global optimum.