Application of Machine Learning Tools in Disaster Management

Application of Machine Learning Tools in Disaster and Emergency Management Background and Aims: Disasters are now an integral part of people's lives. With the advancement of technologies and the emergence of artificial intelligence, we are faced with a multitude of tools that can be utilized for preparedness and effective response to disasters. This article aims to identify existing machine learning tools for disaster preparedness and response. Methods: This is an applied study conducted cross-sectionally using a narrative literature review approach in databases such as PubMed, Scopus, and Google Scholar. The keywords used include (Machine Learning, Classification, Damage, Earthquake, Disaster, Deep Learning, Internet of Things, Warning System, Prediction, Convolutional Neural Network, Artificial Neural Network, Artificial Intelligence, Decision Tree, Random Forest, Reinforcement Learning) without any time restrictions. The research population consisted of all articles available in these databases. The inclusion criteria for selecting articles included relevance to the research objective and availability of full text. Initially, the titles and abstracts of the articles were reviewed, and if selected, the full text was analyzed. The results of the search were examined through content analysis. Results: After applying the keywords in the databases, approximately ۲۰۰۰ articles were obtained. From this number, ۴۰ articles were selected. The results indicated that the applications of machine learning in disasters can be categorized into three main classes (image classification, object detection, and warning and recommendation systems) and several subcategories (Image Classification: damage identification post-disaster, damage assessment post-disaster, monitoring inaccessible areas. Object Detection: automatic feature detection, search and rescue identification of victims post-disaster, minimizing casualties from disasters. Warning Systems: remote sensing technology, early warnings for disasters, prevention strategies). Conclusion: Given the useful capabilities of machine learning in various phases of crisis management, it is essential for health professionals to pay attention to these functionalities and utilize this science to mitigate risks and damages, as well as to ensure a rapid response to disasters to reduce human and financial losses. Keywords: MachineLearning, DeepLearning, Damage, Disaster, Artificial Intelligence

کلیدواژه ها:

MachineLearning ، DeepLearning ، Damage ، Disaster ، Artificial Intelligence