Machine Learning in Agriculture: A Comprehensive Updated Review

Sensors, Journal Year: 2021, Volume and Issue: 21(11), P. 3758 - 3758

Published: May 28, 2021

The digital transformation of agriculture has evolved various aspects management into artificial intelligent systems for the sake making value from ever-increasing data originated numerous sources. A subset intelligence, namely machine learning, a considerable potential to handle challenges in establishment knowledge-based farming systems. present study aims at shedding light on learning by thoroughly reviewing recent scholarly literature based keywords’ combinations “machine learning” along with “crop management”, “water “soil and “livestock accordance PRISMA guidelines. Only journal papers were considered eligible that published within 2018–2020. results indicated this topic pertains different disciplines favour convergence research international level. Furthermore, crop was observed be centre attention. plethora algorithms used, those belonging Artificial Neural Networks being more efficient. In addition, maize wheat as well cattle sheep most investigated crops animals, respectively. Finally, variety sensors, attached satellites unmanned ground aerial vehicles, have been utilized means getting reliable input analyses. It is anticipated will constitute beneficial guide all stakeholders towards enhancing awareness advantages using contributing systematic topic.

Language: Английский

---

Novel ensemble machine learning models in flood susceptibility mapping

Geocarto International, Journal Year: 2021, Volume and Issue: 37(16), P. 4571 - 4593

Published: Feb. 19, 2021

The research aims to propose the new ensemble models by combining machine learning techniques, such as rotation forest (RF), nearest shrunken centroids (NSC), k-nearest neighbour (KNN), boosted regression tree (BRT), and logitboost (LB) with base classifier adabag (AB) for flood susceptibility mapping (FSM). proposed were implemented in central west coast of India, which is vulnerable events. For inventory mapping, a total 210 localities identified. Twelve effective factors selected using boruta algorithm FSM. area under receiver operating characteristics (AUROC) curve other statistical measures (sensitivity, specificity, accuracy, kappa, root mean square error (RMSE), absolute (MAE)) employed estimate compare success rate approaches. validation results individual terms AUC value AB (92.74%) >RF (91.50%) >BRT (90.75%) >LB (89.07%) >NSC (88.97%) >KNN (83.88%), whereas showed that AB-RF (94%) was highest prediction efficiency followed by, AB-KNN (93.33%), AB-NSC (93.02%), AB-LB (92.83%), AB-BRT (92.64%). outcomes established more appropriate increase accuracy different single models. Therefore, this study can be useful proper planning management hazard alike geographic environment.

Language: Английский

---

Forecasting the strength of graphene nanoparticles-reinforced cementitious composites using ensemble learning algorithms

Results in Engineering, Journal Year: 2024, Volume and Issue: 21, P. 101837 - 101837

Published: Feb. 6, 2024

Contemporary infrastructure requires structural elements with enhanced mechanical strength and durability. Integrating nanomaterials into concrete is a promising solution to improve However, the intricacies of such nanoscale cementitious composites are highly complex. Traditional regression models encounter limitations in capturing these intricate compositions provide accurate reliable estimations. This study focuses on developing robust prediction for compressive (CS) graphene nanoparticle-reinforced (GrNCC) through machine learning (ML) algorithms. Three ML models, bagging regressor (BR), decision tree (DT), AdaBoost (AR), were employed predict CS based comprehensive dataset 172 experimental values. Seven input parameters, including graphite nanoparticle (GrN) diameter, water-to-cement ratio (wc), GrN content (GC), ultrasonication (US), sand (SC), curing age (CA), thickness (GT), considered. The trained 70 % data, remaining 30 data was used testing models. Statistical metrics as mean absolute error (MAE), root square (RMSE) correlation coefficient (R) assess predictive accuracy DT AR demonstrated exceptional accuracy, yielding high coefficients 0.983 0.979 training, 0.873 0.822 testing, respectively. Shapley Additive exPlanation (SHAP) analysis highlighted influential role positively impacting CS, while an increased (w/c) negatively affected CS. showcases efficacy techniques accurately predicting nanoparticle-modified concrete, offering swift cost-effective approach assessing nanomaterial impact reducing reliance time-consuming expensive experiments.

Language: Английский

---

River Water Salinity Prediction Using Hybrid Machine Learning Models

Water, Journal Year: 2020, Volume and Issue: 12(10), P. 2951 - 2951

Published: Oct. 21, 2020

Electrical conductivity (EC), one of the most widely used indices for water quality assessment, has been applied to predict salinity Babol-Rood River, greatest source irrigation in northern Iran. This study uses two individual—M5 Prime (M5P) and random forest (RF)—and eight novel hybrid algorithms—bagging-M5P, bagging-RF, subspace (RS)-M5P, RS-RF, committee (RC)-M5P, RC-RF, additive regression (AR)-M5P, AR-RF—to EC. Thirty-six years observations collected by Mazandaran Regional Water Authority were randomly divided into sets: 70% from period 1980 2008 was as model-training data 30% 2009 2016 testing validate models. Several variables—pH, HCO3−, Cl−, SO42−, Na+, Mg2+, Ca2+, river discharge (Q), total dissolved solids (TDS)—were modeling inputs. Using EC correlation coefficients (CC) variables, a set nine input combinations established. TDS, effective variable, had highest EC-CC (r = 0.91), it also determined be important variable among combinations. All models trained each model’s prediction power evaluated with data. quantitative criteria visual comparisons evaluate capabilities. Results indicate that, cases, algorithms enhance individual algorithms’ predictive powers. The AR algorithm enhanced both M5P RF predictions better than bagging, RS, RC. performed RF. Further, AR-M5P outperformed all other (R2 0.995, RMSE 8.90 μs/cm, MAE 6.20 NSE 0.994 PBIAS −0.042). hybridization machine learning methods significantly improved model performance capture maximum values, which is essential resource management.

Language: Английский

---

Comparative study of machine learning models for evaluating groundwater vulnerability to nitrate contamination

Ecotoxicology and Environmental Safety, Journal Year: 2021, Volume and Issue: 229, P. 113061 - 113061

Published: Dec. 11, 2021

The accurate evaluation of groundwater contamination vulnerability is essential for the management and prevention in watershed. In this study, advanced multiple machine learning (ML) models Radial Basis Neural Networks (RBNN), Support Vector Regression (SVR), ensemble Random Forest (RFR) were applied to determine most performance vulnerability. Eight factors DRASTIC-L rated based on modified DRASTIC model (MDM) used as input data. adjusted index (AVI) with nitrate values was output data modeling process. three verified using statistical criteria MAE, RMSE, r2, ROC/AUC values. RFR showed highest comparison standalone SVR RBNN models. Specifically, kept all promising solutions during due its flexibility robustness, map obtained by more predicting vulnerable areas contamination. It concluded that a robust tool enhance vulnerability, it could contribute environmental safety against

Language: Английский

---

Susceptibility mapping of groundwater salinity using machine learning models

Environmental Science and Pollution Research, Journal Year: 2020, Volume and Issue: 28(9), P. 10804 - 10817

Published: Oct. 25, 2020

Language: Английский

---

Comparative Assessment of Individual and Ensemble Machine Learning Models for Efficient Analysis of River Water Quality

Sustainability, Journal Year: 2022, Volume and Issue: 14(3), P. 1183 - 1183

Published: Jan. 21, 2022

The prediction accuracies of machine learning (ML) models may not only be dependent on the input parameters and training dataset, but also whether an ensemble or individual model is selected. present study based comparison supervised ML models, such as gene expression programming (GEP) artificial neural network (ANN), with that model, i.e., random forest (RF), for predicting river water salinity in terms electrical conductivity (EC) dissolved solids (TDS) Upper Indus River basin, Pakistan. projected were trained tested by using a dataset seven chosen basis significant correlation. Optimization RF was achieved producing 20 sub-models order to choose accurate one. goodness-of-fit assessed through well-known statistical indicators, coefficient determination (R2), mean absolute error (MAE), root squared (RMSE), Nash–Sutcliffe efficiency (NSE). results demonstrated strong association between inputs modeling outputs, where R2 value found 0.96, 0.98, 0.92 GEP, RF, ANN respectively. comparative performance proposed methods showed relative superiority compared GEP ANN. Among sub-models, most yielded equal 0.941 0.938, 70 160 numbers corresponding estimators. lowest RMSE values 1.37 3.1 testing data, sensitivity analysis HCO3− effective variable followed Cl− SO42− both EC TDS. assessment external criteria ensured generalized all aforementioned techniques. Conclusively, outcome research indicated selected key could prioritized quality management.

Language: Английский

---

Computational assessment of groundwater salinity distribution within coastal multi-aquifers of Bangladesh

Scientific Reports, Journal Year: 2022, Volume and Issue: 12(1)

Published: July 1, 2022

The rising salinity trend in the country's coastal groundwater has reached an alarming rate due to unplanned use of agriculture and seawater seeping into underground sea-level rise caused by global warming. Therefore, assessing is crucial for status safe aquifers. In this research, a rigorous hybrid neurocomputing approach comprised Adaptive Neuro-Fuzzy Inference System (ANFIS) hybridized with new meta-heuristic optimization algorithm, namely Aquila (AO) Boruta-Random forest feature selection (FS) was developed estimating multi-aquifers regions Bangladesh. regard, 539 data samples, including ten water quality indices, were collected provide predictive model. Moreover, individual ANFIS, Slime Mould Algorithm (SMA), Ant Colony Optimization Continuous Domains (ACOR) coupled ANFIS (i.e., ANFIS-SMA ANFIS-ACOR) LASSO regression (Lasso-Reg) schemes examined compare primary Several goodness-of-fit such as correlation coefficient (R), root mean squared error (RMSE), Kling-Gupta efficiency (KGE) used validate robustness models. Here, Forest (B-RF), robust tree-based FS, adopted identify most significant candidate inputs effective input combinations reduce computational cost time modeling. outcomes four selected ascertained that ANFIS-OA regarding best accuracy terms (R = 0.9450, RMSE 1.1253 ppm, KGE 0.9146) outperformed 0.9406, 1.1534 0.8793), ANFIS-ACOR 0.9402, 1.1388 0.8653), Lasso-Reg 0.9358), 0.9306) Besides, first combination (C1) three inputs, Cl- (mg/l), Mg2+ Na+ yielded among all alternatives, implying role importance (B-RF) selection. Finally, spatial distribution assessment study area high predictability potential B-RF compared other paradigms. important novelty research using framework non-linear filtering technique neuro-computing approach, which can be considered reliable tool assess

Language: Английский

---

Data-driven prediction of neutralizer pH and valve position towards precise control of chemical dosage in a wastewater treatment plant

Journal of Cleaner Production, Journal Year: 2022, Volume and Issue: 348, P. 131360 - 131360

Published: March 16, 2022

Language: Английский

---

Current status and future challenges of groundwater vulnerability assessment: A bibliometric analysis

Journal of Hydrology, Journal Year: 2022, Volume and Issue: 615, P. 128694 - 128694

Published: Nov. 8, 2022

Language: Английский

---