A Machine Learning-based Decision Support System for Disaster Response: A Mobile Approach

Abstract

Machine learning has become one of the most evolving innovations in the field of technology which has a great wide variety of applications. While disaster management, in the hope of the urgent response to the affected communities, the concept of the latter is attributed. The disaster is a serious disturbance in a society that causes human life, physical resources, financial and environmental losses that result from concerns on the ability of the community to handle using its own means and properties. Generally, the objective of this research is to develop a machine learning mobile-based system for disaster management that can respond in a timely manner and provides decision support on evacuation mapping, relief good operations, school allocation, and job assistance. The proponents considered the existing process of the various local government agencies on reaching out to the affected communities. The Agile method was employed on system development to ensure the efficiency and transparency of the actual system design. Based on system and unit testing, it was revealed that the system meets the specified requirements of the various agencies. Hence, the K-Means Clustering was utilized to produce information that enables the organization to efficiently make decisions, validate the results and improve the activities that support the risk reduction management. The study proved that the application of machine learning in Disaster Management can dramatically improve the actions and can manage the data in decision making for future references.