Evaluation and prediction of compound geohazards in highly urbanized regions across China's Greater Bay Area

Journal of Cleaner Production, Journal Year: 2024, Volume and Issue: 449, P. 141641 - 141641

Published: March 16, 2024

Language: Английский

---

Interferometric Synthetic Aperture Radar (InSAR)-Based Absence Sampling for Machine-Learning-Based Landslide Susceptibility Mapping: The Three Gorges Reservoir Area, China

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(13), P. 2394 - 2394

Published: June 29, 2024

The accurate prediction of landslide susceptibility relies on effectively handling absence samples in machine learning (ML) models. However, existing research tends to generate these feature space, posing challenges field validation, or using physics-informed models, thereby limiting their applicability. rapid progress interferometric synthetic aperture radar (InSAR) technology may bridge this gap by offering satellite images with extensive area coverage and precise surface deformation measurements at millimeter scales. Here, we propose an InSAR-based sampling strategy for mapping the Badong–Zigui near Three Gorges Reservoir, China. We achieve employing a Small Baseline Subset (SBAS) InSAR annual average ground deformation. Subsequently, select from slopes very slow Logistic regression, support vector machine, random forest models demonstrate improvement when samples, indicating enhanced accuracy reflecting non-landslide conditions. Furthermore, compare different integration methods integrate into ML including sampling, joint training, overlay weights, combination, finding that utilizing all three simultaneously optimally improves

Language: Английский

---

Towards a Synergistic Progressive Ensemble Framework for Automatic Post-Earthquake Landslide Recognition and Susceptibility Assessment

Mathematical Geosciences, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 3, 2025

Language: Английский

---

Impact of Non-Landslide Sample Sampling Strategies and Model Selection on Landslide Susceptibility Mapping

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(4), P. 2132 - 2132

Published: Feb. 18, 2025

This study investigated the influence of non-landslide sampling strategies on landslide susceptibility assessment (LSA) performance and explored approaches to minimizing uncertainty in model selection. Five were evaluated using random forest (RF) generate maps (LSMs) for each scenario. To assess impact these strategies, this employed a receiver operating characteristic (ROC) curve, confusion matrix, various statistical indicators. Additionally, mean indices derived from gradient boosting decision tree (GBDT), support vector machine (SVM), RF models analyzed evaluate their effectiveness reducing during The GBDT, SVM, selected ability handle complex, nonlinear relationships data, superior generalization capability, effective mitigation overfitting risks, high predictive performance, robustness. findings revealed that selecting samples slope units without landslides enhances accuracy averaging across mitigated associated with models. Furthermore, demonstrated sample selection method significantly improved prediction accuracy, particularly when drawn very-low-susceptibility zones identified by pre-classified learning These results highlight importance refining integrating multiple improve reliability assessments. approach provides valuable insights future research practical applications risk disaster management offering more precise depiction low-susceptibility areas, thereby occurrence false positives prediction.

Language: Английский

---

Effect of different mapping units, spatial resolutions, and machine learning algorithms on landslide susceptibility mapping at the township scale

Environmental Earth Sciences, Journal Year: 2025, Volume and Issue: 84(5)

Published: Feb. 26, 2025

Language: Английский

---