An ensemble random forest tree with SVM, ANN, NBT, and LMT for landslide susceptibility mapping in the Rangit River watershed, India

Summary

This 2022 study develops an ensemble machine learning framework combining five classification algorithms to predict landslide susceptibility across the Rangit River watershed in India's mountainous terrain. The ensemble approach—integrating random forest, SVM, ANN, NBT and LMT—was hypothesised to improve predictive robustness over single-method approaches, with potential application to climate adaptation and disaster risk reduction in flood and landslide-vulnerable watersheds. The work addresses geomorphological hazard assessment in regions vulnerable to extreme precipitation events and terrain instability.

Regional applicability

India-specific study in the Rangit River watershed. Findings may have transferability to similar high-relief, monsoon-influenced watersheds in South Asia and Southeast Asia with comparable geology and precipitation patterns, though UK application would require re-calibration for distinctly different topography, climate regimes, and soil mechanics. Machine learning ensemble methodology itself is globally transferable.

Key measures

Landslide susceptibility predictions; model accuracy metrics; ensemble classifier performance; spatial hazard mapping outputs

Outcomes reported

The study applied an ensemble machine learning approach combining random forest, support vector machine, artificial neural network, naïve Bayes tree, and logistic model tree algorithms to map landslide susceptibility in the Rangit River watershed. The comparative analysis assessed predictive accuracy of the integrated ensemble model against individual algorithm performance.

Topic tags