Landslide Susceptibility Analysis and its Verification usingLikelihood Ratio, Logistic Regression and Artificial NeuralNetwork Methods: Case study of Yongin, Korea

摘要

The likelihood ratio, logistic regression and artificialrnneural networks methods are applied and verified for analysisrnof landslide susceptibility in Yongin, Korea using GIS. From arnspatial database containing such data as landslide location,rntopography, soil, forest, geology and land use, the 14 landsliderelatedrnfactors were calculated or extracted. Using these factors,rnlandslide susceptibility indexes were calculated by likelihoodrnratio, logistic regression and artificial neural network methods.rnBefore the calculation, the study area was divided into twornsides (west and east) of equal area, for verification of the methods.rnThus, the west side was used to assess the landslide susceptibility,rnand the east side was used to verify the derivedrnsusceptibility. The results of the landslide susceptibility analysisrnwere verified using success and prediction rates. The v erificationrnresults showed satisfactory agreement between the susceptibilityrnmap and the exis ting data on landslide locations.