Landslide susceptibility mapping at Zonouz Plain, Iran using genetic programming and comparison with frequency ratio, logistic regression, and artificial neural network models

摘要: Without a doubt, landslide is one of the most disastrous natural hazards and susceptibility maps (LSMs) in regional scale are useful guide to future development planning. Therefore, importance generating LSMs through different methods popular international literature. The goal this study was evaluate occurrence landslides Zonouz Plain, located North-West Iran. For purpose, inventory map constructed using field survey, air photo/satellite image interpretation, literature search for historical records. Then, seven landslide-conditioning factors such as lithology, slope, aspect, elevation, land cover, distance stream, road were utilized generation by various models: frequency ratio (FR), logistic regression (LR), artificial neural network (ANN), genetic programming (GP) geographic information system (GIS). Finally, total four obtained these methods. verification, results LSM analyses confirmed containing 190 active zones. validation process showed that prediction accuracy LSMs, produced FR, LR, ANN, GP, 87.57, 89.42, 92.37, 93.27 %, respectively. indicated use GP provides more accurate comparison with ANN. Furthermore; model superior ANN because it can present an explicit formulation instead weights biases matrices.