Artificial intelligence for remote sensing-based detection and prediction of landslides in Malaysia and Vietnam: A state-of-the-art review

This paper reviews the current status of artificial intelligence (AI) in assessing and predicting landslides, emphasising Malaysia and Vietnam landslides. Based on 95 papers selected through a rigorous literature search and review, it is found that the Support Vector Machine, Dempster-Shafer theory, and adaptive neuro-fuzzy inference systems show better performance for landslide events from Malaysia. In comparison, the tree-based learning algorithms work well for landslide events from Vietnam.

The review outcome also enlightens the application of Geographic Information Systems and remote sensing technology, combined with AI, improves the performance of landslide risk detection and prediction performance. Nevertheless, the implementation of AI is highly variable depending on the quality of data collected from past landslide events, the algorithm’s architecture, and data preprocessing methods. To this end, exploring the variation of landslide risk prediction outcomes, aided by real-time monitoring of slopes with AI, are likely the future in achieving more robust landslide early detection and warning systems. Such systems will ensure safe and reliable monitoring, detection, and prediction of landslides by facilitating timely warnings to society and concerned authorities.