Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 8, 2026SHILAP Revista de lepidopterologíaOpen Access

Sampling effects on machine-learning performance and tectonic controls on landslide susceptibility: insights from the Adra River basin (SE Spain)

View Full Paper
Ask AI
Bookmark
Share

Authors

SBS. BoussoufTFT. FernándezMSMario Sánchez‐Gómez

Discussion

Loading...

Member takes

Overview

Demonstrates improved landslide susceptibility assessment in a diverse terrain, indicating significant geological influences.

Key Points

  • The aim is to develop a better method for assessing landslide susceptibility influenced by geology and tectonics.
  • Compared centroid-based and proportional stratified sampling methods.
  • Tested seven machine-learning models and one mean-ensemble configuration.
  • Reduced landslide conditioning factors from 19 to 14 through feature selection.
  • Evaluated model performance using predictive accuracy and Degree of Fit metrics.
  • The GWLR-Mean-PSS ensemble achieved a high predictive accuracy (AUC = 0.960; Accuracy = 89%).
  • Demonstrated low spatial bias and a Degree of Fit of 96.69%.
  • Susceptibility patterns correspond closely with the basin’s geological and tectonic features.

Cite This Study

Boussouf et al. (2026) studied this question.

synapsesocial.com/papers/698829410fc35cd7a8849678https://doi.org/10.1080/19475705.2026.2620641
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched one closely related paper. Consider it for comparative context:

  1. 1Understanding the fluvial capture of the Guadix-Baza Basin in SE Spain through its oldest exorheic deposits2024 · 6 citations