A Novel GIS-Based Random Forest Machine Algorithm for the Spatial Prediction of Shallow Landslide Susceptibility

Abstract

This study developed and verified a new hybrid machine learning model, named random forest machine (RFM), for the spatial prediction of shallow landslides. RFM is a hybridization of two state-of-the-art machine learning algorithms, random forest classifier (RFC) and support vector machine (SVM), in which RFC is used to generate subsets from training data and SVM is used to build decision functions for these subsets. To construct and verify the hybrid RFM model, a shallow landslide database of the Lang Son area (northern Vietnam) was prepared. The database consisted of 101 shallow landslide polygons and 14 conditioning factors. The relevance of these factors for shallow landslide susceptibility modeling was assessed using the ReliefF method. Experimental results pointed out that the proposed RFM can help to achieve the desired prediction with an F1 score of roughly 0.96. The performance of the RFM was better than those of benchmark approaches, including the SVM, RFC, and logistic regression. Thus, the newly developed RFM is a promising tool to help local authorities in shallow landslide hazard mitigations.

A comparison among fuzzy multi-criteria decision making, bivariate, multivariate and machine learning models in landslide susceptibility mapping

“…Recently, several machine learning algorithms and modelling techniques were used for preparing and assessing LS (Chen et al 2018a(Chen et al , 2018bKavzoglu et al 2018). The ensemble learning methods in LS were prioritized because of their high prediction accuracy (Hong et al 2018;Vakhshoori et al 2019;Dang et al 2020;Tien Bui et al 2020). A recent study compared RFC, SVM and XGBoost for multi-geohazards susceptibility in Jiuzhaigou, China and found that XGBoost had the best prediction capability with highest AUC of 93% (Cao et al 2020).…”

“…Indeed, there are hundreds of recent studies that applied single or ensemble machine learning models. Based on models applied, it is very difficult to find any novelty amongst various studies, but each study is novel in term of geographical location and study findings because different outcomes have been found by applying same models in different geographical location (Achour and Pourghasemi 2020;Cao et al 2020;Dang et al 2020;Tien Bui et al 2020;Ali et al 2021). The main goal of this study was to present a comparative analysis amongst fuzzy MCDM, bivariate statistics, multivariate statistics and machine learning algorithms.…”

“…As a result, in the time of machine learning (ML) application in LS and spatial prediction ( Shirzadi et al 2018;Vakhshoori et al 2019;Achour and Pourghasemi 2020;Cao et al 2020;Dang et al 2020;Tien Bui et al 2020), the purpose of the current study was to compare the performance and prediction accuracy of other techniques also along with machine learning algorithms. The results showed that although all of these models had high rate of prediction accuracy (AUC¼ >80%), but still some heuristic models, particularly fuzzy DEMATEL-ANP can offer better accuracy than selected machine learning techniques, particularly, NBC and XGBoost.…”

Predicting Rainfall-Induced Soil Erosion Based on a Hybridization of Adaptive Differential Evolution and Support Vector Machine Classification

“…To further confirm the superiority of the proposed hybrid L-SHADE-PWI-SVM model, the Wilcoxon signed-rank test [119] with the significant level (p value) � 0.05 is also employed in this study to demonstrate the statistical significance of the difference in model results. is nonparametric hypothesis testing method is widely employed for comparing classification models [120].…”

Stochastic spatial random forest (SS-RF) for interpolating probabilities of missing land cover data

“…Machine learning methods are popular and useful for remote sensing applications including for identifying clouds and cloud shadow as a preprocessing step to interpolation [11,17], and monitoring changes in forest cover [18]. Machine learning algorithms are also popular for important environmental monitoring beyond remote sensing applications including identifying deforestation [19] and landslide susceptibility [20].…”