Gully Erosion Susceptibility Assessment in the Kondoran Watershed Using Machine Learning Algorithms and the Boruta Feature Selection

Abstract

Gully erosion susceptibility mapping is an essential land management tool to reduce soil erosion damages. This study investigates gully susceptibility based on multiple diagnostic analysis, support vector machine and random forest algorithms, and also a combination of these models, namely the ensemble model. Thus, a gully susceptibility map in the Kondoran watershed of Iran was generated by applying these models on the occurrence and non-occurrence points (as the target variable) and several predictors (slope, aspect, elevation, topographic wetness index, drainage density, plan curvature, distance to streams, lithology, soil texture and land use). The Boruta algorithm was used to select the most effective variables in modeling gully erosion susceptibility. The area under the receiver operating characteristic curve (AUC), the receiver operating characteristics, and true skill statistics (TSS) were used to assess the model performance. The results indicated that the ensemble model had the best performance (AUC = 0.982, TSS = 0.93) compared to the others. The most effective factors in gully erosion susceptibility mapping of the study region were topological, anthropogenic, and geological. The methodology and variables of this study can be used in other regions to control and mitigate the gully erosion phenomenon by applying biophilic and regenerative techniques at the locations of the most influential factors.

Gully erosion susceptibility mapping using four machine learning methods in Luzinzi watershed, eastern Democratic Republic of Congo

“…These findings have been supported by many other studies in which a location vegetation and man activities coupled with hydrologic properties are assigned as determinants of the highest gully erosion susceptibility (Imwangana, 2014;Ahmadpour et al, 2021). Other results have suggested that distance to streams and roads, drainage density, and NDVI are significant factors that promote favorable conditions for gullying (Arabameri et al, 2020a).…”

“…When the gain in minimizing the error becomes insignificant, the procedure ends. BRT has shown good predictive performance in predicting gully erosion susceptibility (Ahmadpour et al, 2021;Garosi et al, 2018). In this study, BRT model was fitted using 'Dismo' package (Ridgway et al, 2008), and using settings as recommended by Elith et al (2008).…”

Risk-Profile and Feature Selection Comparison in Diabetic Retinopathy

“…The main idea of this method is to choose important features that are essential for the discovery of new knowledge [27]; a small subset of the significant features subsequently reduce the computation time. Boruta was implemented for this task, the wrapper algorithm determines the importance for each feature; the benefits of this algorithm is that it considers all attributes that are related to the output, and it considers multi-variable relationships where it can also explore interactions between variables [28]; the risk factors for DR were identified with the most important predictors being lipids treatment, glomerular filtration rate, waist hip-ratio, total cholesterol, high density lipoprotein, and creatinine. These six features comprise the new dataset.…”

A Comprehensive Survey on the Process, Methods, Evaluation, and Challenges of Feature Selection

“…By introducing randomness to the system and gathering data from an ensemble of randomized samples, the influence of random fluctuations and correlations can be mitigated [169]. The studies that suggested the boruta algorithm for FS [170], [171], [172], [173]. Adequate selection of features is must to improve accuracy and efficiency of classifier methods.…”