Abstract
Toward Effective Response to Natural Disasters: A Data Science Approach
Highlights
A Lmost all of us have sometimes found ourselves in a situation where a large number of people gathered together in a particular place: spectators at concerts or sports events, students in school premises, commuters in railway and metro stations, employees in large office buildings
We can distinguish three main parts: i) the construction of a network based on the Geographical Information Systems (GIS) data (GisToGraph algorithm), ii) the dynamic flow modeling, and the solution to evacuation planning, and iii) the modeling of the reconstruction planning and its solution built by employing double deep Q-learning network (DDQN) approach
According to the methodology diagram, we have proposed reconstruction planning with the help of Enriched Undirected Graph (EUG) generated by the GisToGraph algorithm, which contains details of the area to be reconstructed as well physical dependencies
Summary
A Lmost all of us have sometimes found ourselves in a situation where a large number of people gathered together in a particular place: spectators at concerts or sports events, students in school premises, commuters in railway and metro stations, employees in large office buildings. Recent examples are the Australian bushfires [1], [39] killing at least 34 people between June 2019 and May 2020; Hurricane Eta, November 2020, killing at least 150 people in Central America [4], [5]; flash floods killing more than 150 people in Afghanistan in August 2020 [3], [22]; earthquake and tsunami killing a total of 117 people in Greece and Turkey in October 2020 [2] (not to mention animals).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
