Landslide Susceptibility Mapping in Gilgit-Baltistan: Leveraging Machine Learning for Risk Assessment

This study examined the challenging issues of landslide detection along Gilgit-Baltistan's Karakoram Highway (KKH), where seismic activity and a dearth of data were significant issues161 characters. The study selected support vector machine (SVM), random forest (RF), and XGBoost algorithms to balance algorithm performance with ease of implementation. Finally, the study assessed the performance of the algorithms by fine-tuning them and optimizing their parameters. This was done by evaluating metrics such as the Area Under the Curve (AUC) and accuracy. Surprisingly, this comprehensive approach resulted in an average accuracy of 90% for all algorithms. These findings provide valuable insights for improving disaster prevention strategies and guiding land-use planning efforts along the KKH highway.