Exploring different approaches for landslide susceptibility zonation mapping in Manipur: a comparative study of AHP, FR, machine learning, and deep learning models

Abstract

The movement of rock, soil, and other debris down a slope or incline is a geological phenomenon known as a landslide. To analyze the landslide susceptibility (LS) in Manipur, the study develops and compares six heterogeneous models, specifically the Analytical Hierarchy Process (AHP), Frequency Ratio (FR), K-Nearest Neighbors (KNN), Random Forest (RF), Support Vector Machine (SVM), and Deep Learning (DL) models were considered. The study found that the DL is the most intriguing model, with a total accuracy of 97.2%, followed by the RF, KNN, SVM, AHP, and FR, with respective accuracy levels of 94.5%, 93.1%, 92.6%, 85.5%, and 76.9%. © 2024 Mapping Sciences Institute, Australia and Geospatial Council of Australia.