SOUTHEAST ASIA, INTEGRATION, MACHINE LEARNING, MULTI-HAZARD, MULTI-CRITERIA ANALYSIS

• Main Author: Viola Chintia, Adria
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/65517
• Institution: Institut Teknologi Bandung

A hazard is a physical event, phenomenon, or human activity that has the potential to damage and cause loss of life, property, and socioeconomic disturbance. Southeast Asia is a region prone to natural disasters in the world. From July 2012 to May 2020, a total of 1,899 disaster events in Southeast Asia (ASEAN) were recorded, affecting more than 147 million people, of which 84,000 people died and were missing and at least suffered a loss of 17 billion dollars. The dangers that occur in ASEAN are dominated by hydrometeorological hazards. The hydrometeorological hazards of landslides, floods, droughts, wildfires, and storms dominate about 80% of ASEAN's annual disasters. This occurs as a result of extreme climate change which continues to increase every year. These hazards cannot be eliminated, but their effects can be minimized. Various efforts have been developed to model ASEAN hazards. Usually, these hazards are modeled separately, but to improve mitigation and resilience management, hazards can be characterized in a complex manner into multi-hazards. Therefore, the purpose of this research is to model multi-hazard using data integration and method integration. This multi-hazard method uses three types of methods, namely machine learning methods, multi-criteria analysis methods, and integration of machine learning methods and multi-criteria analysis. The result of this study is a multi-hazard hazard analysis in Southeast Asia using different methods with a model accuracy of more than 85%. The results of this susceptibility modeling can be used to identify areas that are susceptible to hazards and improve disaster management and resilience.