Comparative analysis of support vector machine and artificial neural network models for soil cation exchange capacity prediction

Jafarzadeh, Ali Asghar; Pal, Mahesh; Servati, Moslem; FazeliFard, M. H.; Ghorbani, M. A.

doi:10.1007/s13762-015-0856-4

Cited by 36 publications

(12 citation statements)

References 52 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other studies use more complex methods such as ANNs, genetic expression programming and support vector networks to estimate CEC from the aforementioned soil properties (e.g. Liao et al ., ; Emamgolizadeh et al ., ; Jafarzadeh et al ., ). Soil specific surface area and the Atterberg limits have also been used to predict CEC (Yukselen‐Aksoy & Kaya, ).…”

Section: Introductionmentioning

confidence: 97%

Rapid estimation of cation exchange capacity from soil water content

Arthur

2017

European J Soil Science

View full text Add to dashboard Cite

Summary Estimates of cation exchange capacity (CEC) for soil or earth materials are vital for several applications in agriculture and geotechnical engineering. The time‐consuming and laborious nature of standard methods of CEC determination prompted the development of regression‐based techniques and more sophisticated methodologies, such as artificial neural networks (ANNs), to estimate CEC from other soil properties (e.g. clay, organic matter and pH). The present study proposes regression relations for estimating CEC from soil water content at different relative humidity (RH) values that range from 10 to 90% and considering sorption hysteresis. For model development, water adsorption and desorption isotherms were measured for 203 soil samples from different parts of the world with a vapour sorption analyser within an RH range from 3 to 93%. Regression models were developed for both adsorption and desorption for nine RH values (10, 20, 30, 40, 50, 60, 70, 80 and 90%). For 35 independent soil samples, the models predicted CEC accurately (root mean squared error (RMSE): adsorption = 2.47 (0.04) cmol(+) kg−1; desorption = 2.52 (0.09) cmol(+) kg−1) with no clear effect of RH value or sorption direction on the accuracy of prediction. The new models were superior to three models from the literature based on clay, organic matter and pH; the smallest RMSE from these models was 4.13 cmol(+) kg−1 for the same 35 soil samples. Furthermore, the water sorption‐based models performed better than literature models based on data mining tools (ANNs, support vector machines) and visible near infrared spectroscopy when standardized RMSE values were compared. Thus, a simple measure of soil water content at a known RH can serve as an accurate substitute for measuring soil CEC. Highlights A simple method to estimate cation exchange capacity (CEC) accurately is needed. Regression models for estimating CEC from soil water contents were developed and tested. Models predicted CEC accurately with no effect of relative humidity or sorption direction. Proposed models performed better than existing models based on the validation dataset.

show abstract

Section: Introductionmentioning

confidence: 97%

Rapid estimation of cation exchange capacity from soil water content

Arthur

2017

European J Soil Science

View full text Add to dashboard Cite

show abstract

“…(2018) demonstrated that a genetic‐based neural network ensemble (GNNE) is outperformed in estimation of daily soil temperatures as compared to the other methods. Though ANN‐based PTFs performed better in few cases (D'Emilio, Aiello, Consoli, Vanella, & Iovino, 2018; Jafarzadeh, Pal, Servati, FazeliFard, & Ghorbani, 2016), but there are several weaknesses associated with ANN such as several coefficients (not easily interpretable) (Schaap et al., 2001), many types of neurons and their connections (Wösten, Pachepsky, & Rawls, 2001), and over‐fitting and over parameterization due to large number of neurons (Hastie, Tibishrani, & Freidman, 2001).…”

Section: Discussionmentioning

confidence: 99%

Prediction of mean weight diameter of soil using machine learning approaches

et al. 2021

View full text Add to dashboard Cite

Even though research shows that aggregate stability and mean weight diameter (MWD) are critical components of soil health, it is not routinely measured. An alternative approach to the physical measurement is to calculate these values based on routinely measured soil parameters. Therefore, the objective was to compare two artificial intelligence (AI)-based machine learning approaches, that is, support vector machine (SVM) and artificial neural network (ANN) models in prediction of soil wet aggregate stability (quantified by MWD). Soil samples (120) from the Indo-Gangetic Alluvium major soil group, that are characterized as Ustifluvents were used in the study. These samples were analyzed for sand, silt, clay, bulk density (BD), organic carbon (OC), and MWD. The correlation coefficient (r) was highest in case of SVM model with a percentage increase of 16.92 and 2.70 when compared with MLR and ANN respectively. The SVM and ANN models showed 6.36 and 2.12% decrease in RMSE in training dataset while a 14.28% decrease was found for SVM in testing dataset when compared to the multi-linear regression (MLR) model. Results showed that ANN with two neurons (building blocks of ANN) in hidden layer had better performance in predicting MWD than MLR, whereas the radial basis Kernel function based SVM was found to be best for training and testing data of MWD. Soil texture, OC, and BD can be used to predict soil structural stability effectively using SVM. However, additional work is needed to confirm these findings with other soils.

show abstract

“…This results is corroborated by Liao et al, (2014), who compared the models performance of multiple stepwise regression, artificial neural network models and SVM for CEC prediction, and attributed their results to a nonlinear relationship between CEC and soil physicochemical properties. In addition, other study (Jafarzadeh et al, 2016) demonstrated that, despite of the ability of SVM to predict CEC in acceptable limits, there is a poor performance in extrapolating the maximum and minimum values of CEC data. Despite this, uncertainties estimated for SVM predictions may not be associated with an incorrect classification.…”

Section: Capacitymentioning

confidence: 98%

A new methodological framework by geophysical sensors combinations associated with machine learning algorithms to understand soil attributes

Mello

Veloso

Lana

et al. 2021

Preprint

View full text Add to dashboard Cite

Abstract. Geophysical sensors combined with machine learning algorithms have been used to understand the pedosphere system, landscape processes and to model soil attributes. In this research, we used parent material, terrain attributes and data from geophysical sensors in different combinations, to test and compare different and novel machine learning algorithms to model soil attributes. Also, we analyzed the importance of pedoenvironmental variables in predictive models. For that, we collected soil physico-chemical and geophysical data (gamma-ray emission from uranium, thorium and potassium, magnetic susceptibility and apparent electric conductivity) by three sensors, gamma-ray spectrometer – RS 230, susceptibilimeter KT10 – Terraplus and Conductivimeter – EM38 Geonics) at 75 points and, we performed soil analysis afterwards. The results showed varying models with the best performance (R2 > 0.2) for clay, sand, Fe2O3, TiO2, SiO2 and Cation Exchange Capacity prediction. Modeling with selection of covariates at three phases (variance close to zero, removal by correction and removal by importance), demonstrated to be adequate to increase the parsimony. The prediction of soil attributes by machine learning algorithms demonstrated adequate values for field collected data, without any sample preparation, for most of the tested predictors (R2 ranging from 0.20 to 0.50). Also, the use of four regression algorithms proved important, since at least one of the predictors used one of the tested algorithms. The performances of the best algorithms for each predictor were higher than the use of a mean value for the entire area comparing the values of Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The best combination of sensors that reached the best model performance to predict soil attributes were gamma-ray spectrometer and susceptibilimeter. The most important variables were parent material, digital elevation model, standardized height and magnetic susceptibility for most predictions. We concluded that soil attributes can be efficiently modelled by geophysical data using machine learning techniques and geophysical sensors combinations. The technique can bring light for future soil mapping with gain of time and environment friendly.

show abstract

Comparative analysis of support vector machine and artificial neural network models for soil cation exchange capacity prediction

Cited by 36 publications

References 52 publications

Rapid estimation of cation exchange capacity from soil water content

Rapid estimation of cation exchange capacity from soil water content

Prediction of mean weight diameter of soil using machine learning approaches

A new methodological framework by geophysical sensors combinations associated with machine learning algorithms to understand soil attributes

Contact Info

Product

Resources

About