Mapping the spatial and temporal variability of flood hazard affected by climate and land-use changes in the future

Journal of Environmental Management, Год журнала: 2021, Номер 298, С. 113551 - 113551

Опубликована: Авг. 17, 2021

The predicts current and future flood risk in the Kalvan watershed of northwestern Markazi Province, Iran. To do this, 512 non-flood locations were identified mapped. Twenty flood-risk factors selected to model using several machine learning techniques: conditional inference random forest (CIRF), gradient boosting (GBM), extreme (XGB) their ensembles. investigate (year 2050) effects changing climates land use on risk, a general circulation (GCM) with representative concentration pathways (RCPs) 2.6 8.5 scenarios by 2050 was tested for impacts 8 precipitation variables. In addition, uses prepared CA-Markov model. performances models validated Receiver Operating Characteristic-Area Under Curve (ROC-AUC) other statistical analyses. AUC value ROC curve indicates that ensemble had highest predictive power (AUC = 0.83) followed GBM 0.80), XGB 0.79), CIRF 0.78). results climate changes flood-prone areas showed classified as having moderate very high will increase 2050. Due occurring climates, area increased predictions from all four models. areal proportion classes zones under RCP scenario have changed following proportions distribution Very Low −12.04 %, −8.56 Moderate +1.56 High +11.55 +7.49 %. has caused present percentages: −14.48 −6.35 +4.54 +10.61 +5.67 mapping can aid planners hazard managers efforts mitigate impacts.

Язык: Английский

---

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

Geomatics Natural Hazards and Risk, Год журнала: 2021, Номер 12(1), С. 1741 - 1777

Опубликована: Янв. 1, 2021

Landslides are dangerous events which threaten both human life and property. The study aims to analyze the landslide susceptibility (LS) in Kysuca river basin, Slovakia. For this reason, previous were analyzed with 16 conditioning factors. Landslide inventory was divided into training (70% of locations) validating dataset (30% locations). heuristic approach Fuzzy Decision Making Trial Evaluation Laboratory (FDEMATEL)-Analytic Network Process (ANP) applied first, followed by bivariate Frequency Ratio (FR), multivariate Logistic Regression (LR), Random Forest Classifier (RFC), Naïve Bayes (NBC) Extreme Gradient Boosting (XGBoost), respectively. results showed that 52.2%, 36.5%, 40.7%, 50.6%, 43.6% 40.3% total basin area had very high LS corresponding FDEMATEL-ANP, FR, LR, RFC, NBC XGBoost model, analysis revealed RFC most accurate model (overall accuracy 98.3% AUC 97.0%). Besides, FDEMATEL-ANP 93.8% 92.4%) better prediction capability than FR 86.9% 86.1%), LR 90.5% 91.2%), machine learning 76.3% 90.9%) even deep 92.3% 87.1%) models. outweighed models, suggests methods should be tested out before directly applying

Язык: Английский

---

A comparative study of prediction of compressive strength of ultra‐high performance concrete using soft computing technique

Structural Concrete, Год журнала: 2023, Номер 24(4), С. 5538 - 5555

Опубликована: Фев. 11, 2023

Abstract Concrete which is the most commercialized construction material and thus it plays a key role in this era of development hence its evolution utmost importance therefore quality concrete to that highly evolved type namely, ultra‐high performance (UHPC) undeniably boon sector. Though, correlations between technical characteristics UHPC composition mixture are complicated, nonlinear, complex characterize using standard statistical techniques. This paper intended use both deep neural network ensemble machine learning algorithms namely gradient boosting, extreme random forest regressor, extra tree voting regressor trained with an 810 collections 15 input variables predict compressive strength. After adjusting regression model, prediction efficiency generalization ability developed models validated number parameters. It was established all employed performed better at forecasting result, accurate followed by boosting.

Язык: Английский

---

Interpretability of simple RNN and GRU deep learning models used to map land susceptibility to gully erosion

The Science of The Total Environment, Год журнала: 2023, Номер 904, С. 166960 - 166960

Опубликована: Сен. 9, 2023

Язык: Английский

---

Empowering sustainability in the built environment: A technological Lens on industry 4.0 Enablers

Technology in Society, Год журнала: 2023, Номер 76, С. 102427 - 102427

Опубликована: Ноя. 23, 2023

Язык: Английский

---

Landslide hazard assessment based on Bayesian optimization–support vector machine in Nanping City, China

Natural Hazards, Год журнала: 2021, Номер 109(1), С. 931 - 948

Опубликована: Июнь 15, 2021

Язык: Английский

---

Evaluation efficiency of hybrid deep learning algorithms with neural network decision tree and boosting methods for predicting groundwater potential

Geocarto International, Год журнала: 2021, Номер 37(19), С. 5564 - 5584

Опубликована: Апрель 23, 2021

Delineation of the groundwater’s potential zones is a growing phenomenon worldwide due to high demand for fresh groundwater. Therefore, identification groundwater an important tool occurrence, protection, and management purposes. More specifically, in arid semi-arid regions, one most natural resources as it supplies water during drought period. The present research study focused on delineation Saveh City, northern part Markazi Province Iran. mapping was prepared using hybrid deep learning machine algorithm boosted tree (BT), artificial neural network (ANN), (DLNN), (DLT), boosting (DB). This carried out by fourteen conditioning factors 349 each springs non-springs points. performance model validated through statistical analysis sensitivity, specificity, positive predictive values (PPV), negative (NPV), receiver operating characteristic (ROC)-area under curve (AUC) analysis. validation result showed that success rate AUC very good DB (0.87–0.99) other models are also i.e. BT (0.81–0.90), ANN (0.77–0.82), DLNN (0.84–0.86), DLT (0.83–0.91). Among several used this altitude, rainfall, distance fault soil types more modeling. Finally, all had efficiency mapping, but recommended use Deep Boost better results future studies. work will be useful planners optimal planning

Язык: Английский

---

Flash-flood hazard susceptibility mapping in Kangsabati River Basin, India

Geocarto International, Год журнала: 2021, Номер 37(23), С. 6713 - 6735

Опубликована: Июль 12, 2021

Flood-susceptibility mapping is an important component of flood risk management to control the effects natural hazards and prevention injury. We used a remote-sensing geographic information system (GIS) platform machine-learning model develop susceptibility map Kangsabati River Basin, India where flash common due monsoon precipitation with short duration high intensity. And in this subtropical region, climate change's impact helps influence distribution rainfall temperature variation. tested three models-particle swarm optimization (PSO), artificial neural network (ANN), deep-leaning (DLNN)-and prepared final classify flood-prone regions study area. Environmental, topographical, hydrological, geological conditions were included models, was selected based on relations between potentiality causative factors multi-collinearity analysis. The results validated evaluated using area under receiver operating characteristic (ROC) curve (AUC), which indicator current state environment value >0.95 implies greater floods. AUC values for ANN, DLNN, PSO training datasets 0.914, 0.920, 0.942, respectively. Among these showed best performance 0.942. approach applicable eastern part India, allow mitigation help improve region.

Язык: Английский

---

GIS-based comparative study of Bayes network, Hoeffding tree and logistic model tree for landslide susceptibility modeling

CATENA, Год журнала: 2021, Номер 203, С. 105344 - 105344

Опубликована: Апрель 22, 2021

Язык: Английский

---

Pattern Recognition for Distributed Optical Fiber Vibration Sensing: A Review

IEEE Sensors Journal, Год журнала: 2021, Номер 21(10), С. 11983 - 11998

Опубликована: Март 15, 2021

In recent years, pattern recognition technologies for distributed optical fiber vibration sensing have attracted more and attention, aiming to intelligently recognize events along with the fiber. Firstly, sensors detection are introduced. Secondly, this paper presents state of art models used in systems. The feature extraction method, model structure, processing performance reported. As results comparison, support vector machine is a small sample learning method solid theoretical foundation it has excellent generalization ability. artificial neural network suitable massive data multi-classification problems. Also, deep can learn features information by nonlinear structure an automated way, thus better accuracy robustness. Furthermore, different applications provided, including perimeter security, pipeline monitoring, railway safety monitoring. Finally, prospects discussed.

Язык: Английский

---