Scienceline Publications Journal of Civil Engineering and Urbanism 2252-0430 6 6 2016 November Modeling of Urban Vulnerability Probability (Supposition) Using Logistic Regression Model; Case study: Tehran city 94 100 http://www.ojceu.ir/main/attachments/article/55/J.%20Civil%20Eng.%20Urban.,%206%20(6)%2094-100,%202016.pdf EN Saeed Maleki PhD, Associate Professor, Geography and Urban Planning, Faculty of Earth Sciences and GIS, Shahid Chamran University of Ahvaz, Iran Abolfazl Meshkini Associate Professor, Department of Geography and Urban Planning, Tarbiat Modares University, Iran Mohammad Ebrahimi MA student of Geography and Urban Planning, Tarbiat Modares University, Iran Abdolah Kloraghan MA student of GIS, Tarbiat Modares University, Iran Atena Moin Mehr MA student of Geography, Tehran University, Iran Morteza Omidipoor PhD student of GIS, Tehran University, Iran Surveying about city seismic vulnerability is one of the urban management requirements. Using an appropriate model with different spatial and non-spatial data and conducting relevant analysis in Geographic Information System and Multi Criteria Decision model can be as useful tools in urban management. The objective of this study is assessing the vulnerability potential of the region 5 of Tehran using logistic regression model. In this study, 14 indicators are used to estimate the vulnerability: Substructure's area of all floors in the block, distance from crisis management centers, Block's area, distance from Gas reducer center and Transmission substation, block's population, Access to communication road, old texture regions, distance from parks and green space, distance from gas and gasoline stations, distance from fault, distance from fire station centers, distance from hospitals and health centers, Average of the block's floor, Slope of the region. Output from 0.83 to 1 class of the SAW model is used as dependent variable to generate the vulnerability distribution map for Logistic regression analysis in IDRISI software. As a result, blocks population (with value of 37.05) and average of the block's floor (with coefficient of 11.6) have the greatest impact on vulnerability. Roc criterion is used for the evaluation of model. The value for this criterion is calculated as 0.996 which shows strong relationship with the values obtained from the logistic regression model. Vulnerability Logistic Regression Geographic Information System spatial and non-spatial data