Modelling of infiltration of sandy soil using gaussian process regression

Sihag, Parveen; Tiwari, N. K.; Ranjan, Subodh

doi:10.1007/s40808-017-0357-1

Cited by 85 publications

(23 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A lot of studies in the literature discuss the use of modeling techniques, viz. support vector machines (SVM), artificial neural network (ANN), multivariate regression function-based relationships (MLR and MNLR), gene expression programming (GEP), Gaussian process regression (GP) (Onen 2014;Sihag et al 2017Sihag et al , 2018aSingh et al 2017;Kumar et al 2018). In this study, relationships based on multiple linear and nonlinear regression are compared with the back propagation artificial neural network technique.…”

Section: Modelingmentioning

confidence: 99%

Oxygen transfer study and modeling of plunging hollow jets

2018

Self Cite

View full text Add to dashboard Cite

Experiments were performed to study the effect of jet angle of hollow jet aerators on oxygen transfer of water. Three jet angles of 30°, 45° and 60° were used to produce hollow jets impingement on the water in the pool. Results from experimental study showed a significant effect of hollow jet angle on volumetric oxygen transfer coefficient (K L a) at higher jet velocities. Standard oxygen transfer efficiency (SOTE) of 60° hollow jet aerators is observed higher with low-kinetic-power jets, but at higher kinetic power, 30° hollow jet aerator is observed to be the best in terms of SOTE. Moreover, modeling performance of the experimental data is evaluated by training and testing of the models, and empirical equations derived from multiple linear and multiple nonlinear regression techniques are compared with artificial neural network approach. The results of the present study are also compared with the previous studies of plunging jet aerators.

show abstract

Section: Modelingmentioning

confidence: 99%

Oxygen transfer study and modeling of plunging hollow jets

2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…Tiwari et al (2017) used the generalised regression neural network, MLR, M5P model tree and SVM to predict the cumulative infiltration of soil and found that SVM works well than the other techniques. Various researchers have been used various soft computing techniques in hydraulics and environmental engineering applications (Sihag et al 2017b(Sihag et al , c, 2018aHaghiabi et al 2018;Nain et al 2018a;Tiwari et al 2018;Parsaie et al 2017a, b;Shiri et al 2016Shiri et al , 2017Parsaie andHaghiabi 2015, 2017;Parsaie 2016;Azamathulla et al 2016;Baba et al, 2013). These researchers found that these techniques work exceptionally well.…”

Section: Introductionmentioning

confidence: 99%

Modelling of the impact of water quality on the infiltration rate of the soil

2019

Self Cite

View full text Add to dashboard Cite

The concept behind of this paper is to check the potential of the three regression-based techniques, i.e. M5P tree, support vector machine (SVM) and Gaussian process (GP), to estimate the infiltration rate of the soil and to compare with two empirical models, i.e. Kostiakov model and multi-linear regression (MLR). Totally, 132 observations were obtained from the laboratory experiments, out of which 92 observations were used for training and residual 40 for testing the models. A double-ring infiltrometer was used for experimentation with different concentrations (1%, 5%, 10% and 15%) of impurities and different types of water quality (ash and organic manure). Cumulative time (T f ), type of impurities (I t ), concentration of impurities (C i ) and moisture content (W c ) were the input variables, whereas infiltration rate was considered as target. For SVM and GP regression, two kernel functions (radial based kernel and Pearson VII kernel function) were used. The results from this investigation suggest that M5P tree technique is more precise as compared to the GP, SVR, MLR approach and Kostiakov model. Among GP, SVR, MLR approach and Kostiakov model, MLR is more accurate for estimating the infiltration rate of the soil. Thus, M5P tree is a technique which could be used for modelling the infiltration rate for the given data set. Sensitivity analyses suggest that the cumulative time (T f ) is the major influencing parameter on which infiltration rate of the soil depends.

show abstract

“…On the other hand, nonparametric methods or memory‐based learning algorithms do not require that knowledge, making them an interesting option due to the complexity of relationships among soil properties. Many data‐mining‐based PTFs developed to predict soil hydraulic conductivity use artificial neural networks combined with fuzzy logic (e.g., Iversen et al, 2011; Merdun et al, 2006; Arshad et al, 2013; Sihag et al, 2017a). Support vector machines with strong pattern recognition overcome local minima, which are conventional problems of neural networks in the training process (Lamorski et al, 2008; Haghverdi et al, 2014).…”

mentioning

confidence: 99%

Evaluation of Parametric and Nonparametric Machine‐Learning Techniques for Prediction of Saturated and Near‐Saturated Hydraulic Conductivity

Kotlar

Iversen²,

Lier

2019

Vadose Zone Journal

View full text Add to dashboard Cite

Core Ideas Accurate Ks and K10 were obtained using water content data at several matric potentials. Robust Ks and K10 prediction by machine‐learning methods confirmed by bootstrapping. Gaussian process regression predicted Ks and K10 with minimum number of predictors. Parametric and nonparametric supervised machine learning techniques were used to estimate saturated and near‐saturated hydraulic conductivities (Ks and K10, respectively) from easily measurable soil properties including the name of the pedological horizon (HOR), soil texture (sand, silt, and clay), organic matter (OM), bulk density (BD), and water contents (θpF1, θpF2, θpF3, and θpF4.2) measured at four different matric heads (−10, −100, −1000, and −15,848 cm, respectively). Using a stepwise linear model (SWLM) and the Lasso regression as parametric methods with 316 data in training and 135 data in the testing phase, four pedotransfer functions (PTFs) were obtained in which water contents for both methods play an important role compared with other variables. The SWLM showed better performance than Lasso in the testing phase for log(Ks) and log(K10) prediction, with RMSE values of 0.666 and 0.551 cm d−1 and R2 of 0.26 and 0.65. Nonparametric supervised machine learning methods trained and tested with a similar data set significantly improved the accuracy of Ks prediction, with R2 of 0.52, 0.36, and 0.53 for Gaussian process regression (GPR), support vector machine (SVM), and ensemble (ENS) methods in the testing stage. These methods also described 74.9, 66.7, and 72.5% of the variation of log(K10). Bootstrapping validated the strong performance of nonparametric techniques. The feature selection capability of GPR determined that instead of using a model with all predictors, HOR, silt, θpF1, and θpF3 are sufficient for the prediction of log(Ks), while HOR, silt, and OM can predict log(K10) as accurate as the comprehensive model with all variables.

show abstract

Modelling of infiltration of sandy soil using gaussian process regression

Cited by 85 publications

References 24 publications

Oxygen transfer study and modeling of plunging hollow jets

Oxygen transfer study and modeling of plunging hollow jets

Modelling of the impact of water quality on the infiltration rate of the soil

Evaluation of Parametric and Nonparametric Machine‐Learning Techniques for Prediction of Saturated and Near‐Saturated Hydraulic Conductivity

Contact Info

Product

Resources

About