Review on Predicting the Landslide Susceptibility using Machine Learning Techniques

This Landslide susceptibility prediction is a critical task for Alleviating the risks associated with landslide hazards, which have significant socio-economic and environmental impacts. Existing techniques face challenges related to data quality, model coherence, and generalizability across diverse geographical and climatic regions. Additionally, while advanced machine learning (ML) and deep learning (DL) models offer high accuracy, they often require significant computational resources, which limits their practical application in real-time scenarios. In this review, we aim to highlight the strengths, limitations, and potential improvements of machine learning techniques in predicting landslide susceptibility, providing a comprehensive reference for researchers and practitioners in this domain. Furthermore, the role of Explainable AI is examined to address the critical need for interpretable and transparent models in real-world applications. Hybrid ML-DL models outperform traditional methods in handling complex spatial-temporal data, with dynamic factors improving accuracy and reliability. Explainable AI enhances interpretability, while advancements and collaborations address data and generalizability challenges.