La survenue des catastrophes naturelles : classification des variables explicatives par les réseaux de neurones
Keywords:
Natural Disasters, Urban Economics, Risk Management, Neural Network, Sensitivity AnalysisAbstract
In recent decades, the occurrence of natural disasters has been sharply increasing. These disastrous risks hit different countries in many dispersed areas and will most likely continue to be real threats in the world. Since no country is immune to natural disasters. It therefore seemed appropriate to test the determinants of their occurrence particularly with the restriction in their period of return and their increasing chance of occurrence. We booked this paper to test the factors underlying the occurrence of natural disasters. Our work is based on the application of an artificial neural network that is the multilayer perceptron. We use this network to predict the number of natural disasters from the variables the most mentioned by literature. Then, we applied the sensitivity analysis for classifying explanatory variables according to their influence on the natural disasters number during the period of 1990-2007.
The obtained results showed that our neural network predicts well the number of natural disasters. Indeed, all the explanatory variables have a significant effect on the neural output. So, they contribute considerably in the explanation of a complex problem such the occurrence of natural disasters.
Metrics
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
CC BY-NC-ND 4.0 Creative Commons Attribution-NonCommercial-NoDerivative 4.0 Internal licenses. This license allows you to share, copy, distribute and transmit the work for non-commercial purposes, providing attribution is made to the authors (but not in a way that suggests that he endorses you or your use of the work). In order to access detailed and updated information on the license, please visit: https://creativecommons.org/licenses/by-nc-sa/4.0/