Machine learning for predicting natural disasters
- Submission status
- Closed
- Submission deadline
Acute events of natural (atmospheric, biological, geophysical, hydrologic, or oceanographic) origin constitute a persistent threat to our society and natural habitat, which remains highly unpredictable. Such events, which are often referred to as natural disasters, not only affect approximately 160 million people worldwide every year, but also can impact certain regions and populations disproportionately frequently, which further increases the intercommunal capacity gaps.
In recent years, and in the face of a novel acute necessity brought about by a global pandemic, interest has grown in employing machine learning methods in risk information and early warning systems, to bolster natural disaster prediction. In many fields, from medicine to finance, machine learning has gained traction due to shrinking the processing timelines, providing growth in computational power, and increasing the usability of large data sets. In the context of predicting natural disasters, machine learning carries a significant potential for a societal and natural benefit, as it can allow for capitalising on a wealth of existing geospatial data, improving the effectiveness of emergency communications, and increasing the accuracy of forecasts.
This Collection aims to bring together the latest machine learning research which can be applied to the specific problem of improving the prediction of natural disasters, from predictive analysis techniques, to data mining, to disaster risk modelling. This Collection welcomes both theoretical and experimental work.
Editors
-
Sultan Kocaman
Hacettepe University, Turkey
-
Neelima Satyam
Indian Institute of Technology Indore, India
-
Manzhu Yu
The Pennsylvania State University, USA
