A comparison among fuzzy multi-criteria decision making, bivariate, multivariate and machine learning models in landslide susceptibility mapping

Landslides are dangerous events which threaten both human life and property. The study aims to analyze the landslide susceptibility (LS) in Kysuca river basin, Slovakia. For this reason, previous were analyzed with 16 conditioning factors. Landslide inventory was divided into training (70% of locations) validating dataset (30% locations). heuristic approach Fuzzy Decision Making Trial Evaluation Laboratory (FDEMATEL)-Analytic Network Process (ANP) applied first, followed by bivariate Frequency Ratio (FR), multivariate Logistic Regression (LR), Random Forest Classifier (RFC), Naïve Bayes (NBC) Extreme Gradient Boosting (XGBoost), respectively. results showed that 52.2%, 36.5%, 40.7%, 50.6%, 43.6% 40.3% total basin area had very high LS corresponding FDEMATEL-ANP, FR, LR, RFC, NBC XGBoost model, analysis revealed RFC most accurate model (overall accuracy 98.3% AUC 97.0%). Besides, FDEMATEL-ANP 93.8% 92.4%) better prediction capability than FR 86.9% 86.1%), LR 90.5% 91.2%), machine learning 76.3% 90.9%) even deep 92.3% 87.1%) models. outweighed models, suggests methods should be tested out before directly applying