<![CDATA[Background: Landslides are among the most frequent and destructive natural hazards in Indonesia, triggered by complex interactions between topographic conditions and rainfall intensity. Accurate predictive models are essential to support early warning systems and disaster mitigation planning. Objective: This study aims to integrate Geographic Information Systems (GIS) with machine learning techniques to improve the accuracy of landslide risk prediction based on topographic and rainfall variables. Method: Topographic parameters derived from Digital Elevation Model (DEM) data include slope, elevation, curvature, and Topographic Wetness Index (TWI), while rainfall variables consist of daily and cumulative precipitation. Four classification algorithms Logistic Regression, Support Vector Machine (SVM), Random Forest, and XGBoost were trained using an 80:20 data split and validated through k-fold cross-validation. Model performance was evaluated using accuracy, precision, recall, F1-score, and Area Under Curve (AUC). Results: XGBoost achieved the highest performance with an AUC of 0.94 and accuracy of 0.90, outperforming other models. Feature importance analysis indicates that slope and five-day cumulative rainfall are dominant predictors. Overall: The integration of GIS and ensemble learning provides a robust and spatially consistent framework for landslide risk mapping and mitigation planning.]]>
Copyrights © 2024
