| Exploitation of MCDM to learn the radial base neural network (RBFNN) for analyzing physical and social vulnerability versus to earthquakes (Case study: Sanandaj City, Iran) |
| Paper ID : 1404-SMPR |
| Authors: |
|
peyman - yarian *1, Mohammad Reza Karami2 1Department of Geography Information System,Mamaghan branch,Islamic Azad University,Mamaghan,Iran 2Department of Social Science, Faculty of Humanities and Social Sciences, Payam Noor University,Tehran |
| Abstract: |
| Despite years of research on natural hazards and efforts to reduce physical and psychological damage, the earthquake as a natural disaster is a catastrophic.Given that human is the main axis in dealing with crisis and vulnerability and the city's space encompasses the largest population spectrum,manage this space is essential.To this end, Sanandaj city vulnerability will be defined by environmental, physical and social criteria.In this regard, with the aim of modeling and assessing the risk and vulnerability, the MCDM-ANN hybrid model was introduced as a new method for teaching learning models. Firstly, in order to determine the final value of each of the criteria, and to apply the conditions for ANN training, analyze the dynamics AHP as one of the MCDM methods to solve complex and non-structural problems by creating a functional hierarchy it has been used. Then,for mathematical modeling in Terrset's Geotechnical Surveillance and Surveillance Software,of the radial base functional neural network (RBFNN) was used as a radial base function and one of the techniques of artificial neural networks.After the modeling, the maps of AHP and RBFNN were compared and the social and environmental vulnerability,indicating that the vulnerability is higher in areas with a high population density. These neighborhoods Mostly in areas 1 and 2, physical factors have a greater effect on their vulnerability.But in vulnerable neighborhoods in Zone 3, environmental factors are more effective and The validation of the model was done using the ORC curve, which indicates that the model's accuracy was high in assessing the seismic vulnerability. |
| Keywords: |
| Earthquake, MCDM,Machine learning, ANN, RBFNN, Sanandaj, IRAN |
| Status : Conditional Accept (Poster) |
