Novel Ensemble Approach of Deep Learning Neural Network (DLNN) Model and Particle Swarm Optimization (PSO) Algorithm for Prediction of Gully Erosion Susceptibility

Sensors, Journal Year: 2020, Volume and Issue: 20(19), P. 5609 - 5609

Published: Sept. 30, 2020

This study aims to evaluate a new approach in modeling gully erosion susceptibility (GES) based on deep learning neural network (DLNN) model and an ensemble particle swarm optimization (PSO) algorithm with DLNN (PSO-DLNN), comparing these approaches common artificial (ANN) support vector machine (SVM) models Shirahan watershed, Iran. For this purpose, 13 independent variables affecting GES the area, namely, altitude, slope, aspect, plan curvature, profile drainage density, distance from river, land use, soil, lithology, rainfall, stream power index (SPI), topographic wetness (TWI), were prepared. A total of 132 locations identified during field visits. To implement proposed model, dataset was divided into two categories training (70%) testing (30%). The results indicate that area under curve (AUC) value receiver operating characteristic (ROC) considering datasets PSO-DLNN is 0.89, which indicates superb accuracy. rest are associated optimal accuracy have similar model; AUC values ROC DLNN, SVM, ANN for 0.87, 0.85, 0.84, respectively. efficiency terms prediction increased. Therefore, it can be concluded its PSO used as novel practical method predict susceptibility, help planners managers manage reduce risk phenomenon.

Language: Английский

---

Evaluation of different boosting ensemble machine learning models and novel deep learning and boosting framework for head-cut gully erosion susceptibility

Journal of Environmental Management, Journal Year: 2021, Volume and Issue: 284, P. 112015 - 112015

Published: Jan. 27, 2021

Language: Английский

---

Soil erosion potential hotspot zone identification using machine learning and statistical approaches in eastern India

Natural Hazards, Journal Year: 2020, Volume and Issue: 104(2), P. 1259 - 1294

Published: Aug. 6, 2020

Language: Английский

---

Changing climate and land use of 21st century influences soil erosion in India

Gondwana Research, Journal Year: 2021, Volume and Issue: 94, P. 164 - 185

Published: March 13, 2021

Language: Английский

---

Novel Machine Learning Approaches for Modelling the Gully Erosion Susceptibility

Remote Sensing, Journal Year: 2020, Volume and Issue: 12(17), P. 2833 - 2833

Published: Sept. 1, 2020

The extreme form of land degradation caused by the formation gullies is a major challenge for sustainability resources. This problem more vulnerable in arid and semi-arid environment associated damage to agriculture allied economic activities. Appropriate modeling such erosion therefore needed with optimum accuracy estimating regions taking appropriate initiatives. Golestan Dam has faced an acute gully over last decade adversely affected society. Here, artificial neural network (ANN), general linear model (GLM), maximum entropy (MaxEnt), support vector machine (SVM) learning algorithm 90/10, 80/20, 70/30, 60/40, 50/50 random partitioning training validation samples was selected purposively susceptibility. main objective this work predict susceptible zone possible accuracy. For purpose, approaches were implemented. 20 conditioning factors considered predicting areas considering multi-collinearity test. variance inflation factor (VIF) tolerance (TOL) limit assessment reducing error models increase efficiency outcome. ANN sample most optimal analysis. area under curve (AUC) values receiver operating characteristics (ROC) (50/50) data are 0.918 0.868, respectively. importance causative estimated help Jackknife test, which reveals that important topography position index (TPI). Apart from this, prioritization all predicted into account set, should future researchers select perspective. type outcome planners local stakeholders implement water conservation measures.

Language: Английский

---

Ensemble of Machine-Learning Methods for Predicting Gully Erosion Susceptibility

Remote Sensing, Journal Year: 2020, Volume and Issue: 12(22), P. 3675 - 3675

Published: Nov. 10, 2020

Gully formation through water-induced soil erosion and related to devastating land degradation is often a quasi-normal threat human life, as it responsible for huge loss of surface soil. Therefore, gully susceptibility (GES) mapping necessary in order reduce the adverse effect diminishes this type harmful consequences. The principle goal present research study develop GES maps Garhbeta I Community Development (C.D.) Block; West Bengal, India, by using machine learning algorithm (MLA) boosted regression tree (BRT), bagging ensemble BRT-bagging with K-fold cross validation (CV) resampling techniques. combination aforementioned MLAs approaches state-of-the-art soft computing, not used evaluation. In further progress our work, here we total 20 conditioning factors (GECFs) 199 head cut points modelling GES. variables’ importance, which erosion, was determined based on random forest (RF) among several GECFs study. output result model’s performance validated receiver operating characteristics-area under curve (ROC-AUC), sensitivity, specificity, positive predictive value (PPV) negative (NPV) statistical analysis. predicted shows that most well fitted where AUC K-3 fold 0.972, whereas PPV NPV 0.94, 0.93, 0.96 respectively, training dataset, followed BRT model. Thus, from concluded BRT-Bagging can be applied new approach studies spatial prediction outcome work helpful policy makers implementing remedial measures minimize damages caused erosion.

Language: Английский

---

Prediction of gully erosion susceptibility mapping using novel ensemble machine learning algorithms

Geomatics Natural Hazards and Risk, Journal Year: 2021, Volume and Issue: 12(1), P. 469 - 498

Published: Jan. 1, 2021

Spatial modelling of gully erosion at regional level is very relevant for local authorities to establish successful counter-measures and change land-use planning. This work exploring researching the potential a genetic algorithm-extreme gradient boosting (GE-XGBoost) hybrid computer education solution spatial mapping susceptibility erosion. The new machine learning approach combine extreme (XGBoost) algorithm (GA). GA metaheuristic being used improve efficiency XGBoost classification approach. A GIS database has been developed that contains recorded instances incidents 18 conditioning variables. These parameters are as predictive variables assess condition non-erosion or in given region within Kohpayeh-Sagzi River Watershed research area Iran. Exploratory results indicate proposed GE-XGBoost model superior other benchmark with desired precision (89.56%). Therefore, newly built may be promising method large-scale susceptibility.

Language: Английский

---

Evaluation of different DEMs for gully erosion susceptibility mapping using in-situ field measurement and validation

Ecological Informatics, Journal Year: 2021, Volume and Issue: 65, P. 101425 - 101425

Published: Sept. 11, 2021

Language: Английский

---

Optimization modelling to establish false measures implemented with ex-situ plant species to control gully erosion in a monsoon-dominated region with novel in-situ measurements

Journal of Environmental Management, Journal Year: 2021, Volume and Issue: 287, P. 112284 - 112284

Published: March 9, 2021

Language: Английский

---

Flood susceptibility mapping using meta-heuristic algorithms

Geomatics Natural Hazards and Risk, Journal Year: 2022, Volume and Issue: 13(1), P. 949 - 974

Published: April 11, 2022

Flood is a common global natural hazard, and detailed flood susceptibility maps for specific watersheds are important management measures. We compute the map Kaiser watershed in Iran using machine learning models such as support vector (SVM), Particle swarm optimization (PSO), genetic algorithm (GA) along with ensembles (PSO-GA SVM-GA). The application of assessment mapping analyzed, future research suggestions presented. model was constructed based on fifteen causatives: slope, slope aspect, elevation, plan curvature, land use, cover, normalize differences vegetation index (NDVI), convergence (CI), topographical wetness (TWI), topographic positioning Index (TPI), drainage density (DD), distance to stream, terrain ruggedness (TRI), surface texture (TST), geology stream power (SPI) inventory data which later divided by 70% training 30% validated model. output evaluated through sensitivity, specificity, accuracy, precision, Cohen Kappa, F-score, receiver operating curve (ROC). evaluation method shows robust results from (0.839), particle (0.851), (0.874), SVM-GA (0.886), PSO-GA (0.902). Compared have done some methods commonly used this assessment. A high-quality, informative database essential classification types that very helpful improve performances. performance ensemble better than model, yielding high degree accuracy (AUC-0.902%). Our approach, therefore, provides novel studies other watersheds.

Language: Английский

---