Visible and near infrared spectroscopy coupled to random forest to quantify some soil quality parameters

Santana, Felipe Bachion de; Souza, Alexandre Lima de; Poppi, Ronei J.

doi:10.1016/j.saa.2017.10.052

Cited by 88 publications

(50 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Several multivariate models based on VIS-NIR-SWIR have been applied to processing soil spectra in order to mathematically extract meaningful information from individual spectrum to accurately predict chemical and physical soil properties, such as organic carbon/matter, pH, total nitrogen, soil moisture, and cation exchange capacity, among others (Morellos et al, 2016;Demattê et al, 2017;Dotto et al, 2018;Xu et al, 2018). The capacity to predict sand, silt, and clay has also been demonstrated in previous studies (Vendrame et al, 2012;Demattê et al, 2016b;Lacerda et al, 2016;Nawar et al, 2016;Dotto et al, 2017;Santana et al, 2018), but none of them in a regional soil legacy spectral library of subtropical soils in Brazil. Among the multivariate model, the partial least square regression (PLS) is the most common multivariate model used (Dotto et al, 2018), given its simplicity and robustness (Viscarra Rossel et al, 2006;Vasques et al, 2008;Lacerda et al, 2016).…”

Section: Introductionmentioning

confidence: 87%

“…Among the multivariate model, the partial least square regression (PLS) is the most common multivariate model used (Dotto et al, 2018), given its simplicity and robustness (Viscarra Rossel et al, 2006;Vasques et al, 2008;Lacerda et al, 2016). However, other studies have established that nonlinear data-mining models such as Support Vector Machines (SVM), Gaussian Process Regression (GPR), and Random Forest (RF) can outperform PLS when used to build predictive models from reflectance spectra (Terra et al, 2015;Nawar et al, 2016;Dotto et al, 2017;Santana et al, 2018). In addition to these models, another data-mining tool based on Cubist regression-rules has been introduced into the spectroscopy approach to predicting soil properties (Minasny and Mcbratney, 2008;Viscarra Rossel and Webster, 2012;Morellos et al, 2016;Viscarra Rossel et al, 2016;Zeng et al, 2017;Sorenson et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Regional Legacy Soil Dataset for Prediction of Sand and Clay Content with Vis-Nir-Swir, in Southern Brazil

Silva

Giasson

Dotto

et al. 2019

Rev. Bras. Ciênc. Solo

View full text Add to dashboard Cite

The success of soil prediction by VIS-NIR-SWIR spectroscopy has led to considerable investment in large soil spectral libraries. The aims of this study were 1) to develop a soil VIS-NIR-SWIR spectroscopy approach using legacy soil samples to improve spectral soil information in a regional scale; (2) to compare six spectral preprocessing techniques; and (3) to compare the performance of linear and non-linear multivariate models for prediction of sand and clay content. A total of 1,534 legacy soil samples, stored by Epagri, were collected from agricultural areas in 2009 on a regional scale, covering 260 municipalities of Santa Catarina. Six spectral preprocessing techniques were applied and compared with reflectance spectra (control treatment) in the development of sand and clay prediction models. Five multivariate regression models, Support Vector Machines, Gaussian Process Regression, Cubist, Random Forest, and Partial Least Square Regression were compared. The scatter-corrective preprocessing groups produced similar or better performance than spectral-derivatives. In addition, preprocessing spectra prior to regression analysis does not improve sand prediction, since reflectance spectra achieved the best performance using Cubist, SVM, and PLS models. In general, clay content presented better prediction accuracy than sand content. The best multivariate model to predict sand and clay content from soil VIS-NIR-SWIR spectra was Cubist. The best Cubist performance was achieved combined with reflectance spectra (R 2 = 0.73; root mean square error = 10.60 %; ratio of the performance to the interquartile range = 2.36) and MSC (R 2 = 0.83; root mean square error = 7.29 %; ratio of the performance to the interquartile range = 3.70) for sand and clay content, respectively. Considering the mean RMSE values of the validation set, the predictive ability of the multivariate models decreased in the following order: Cubist>PLS>RF>GPR>SVM for both properties. The predictive ability of VIS-NIR-SWIR reflectance spectroscopy achieved in this study for sand and clay content using legacy soil data and heterogeneous samples confirmed the potential of the spectroscopy approach.

show abstract

Section: Introductionmentioning

confidence: 87%

Section: Introductionmentioning

confidence: 99%

A Regional Legacy Soil Dataset for Prediction of Sand and Clay Content with Vis-Nir-Swir, in Southern Brazil

Silva

Giasson

Dotto

et al. 2019

Rev. Bras. Ciênc. Solo

View full text Add to dashboard Cite

show abstract

“…Though most studies remove extreme values from the modelling procedure, there is a notion that they should not be excluded from the dataset for a robust model to be able to detect such values [36]. An example is the work of De Santana et al that aimed at providing adequate information about the methodology used for the evaluation of outliers [37], and proposed a new method for their detection using the random forest (RF) technique. It was found that RF generated a lower number of outliers from partial least square regression (PLSR).…”

Section: Preprocessing Techniquesmentioning

confidence: 99%

“…The study also showed a high level of correlation of SOM and raw spectra between the 550-680 nm regions, probably related to soil color. Due to the complex relationship between spectral data and soil parameters that is not always considered linear, de Santana et al [37] compared RF to PLSR; the results were marginally better for RF, as they were able to identify outliers using a proximity matrix. In the same line, SVMR was found to perform better from linear multivariate methods (principal component analysis-PCA, PLSR) and back-propagation neural networks [51].…”

Section: Multivariate Calibrationsmentioning

confidence: 99%

From Laboratory to Proximal Sensing Spectroscopy for Soil Organic Carbon Estimation—A Review

et al. 2020

View full text Add to dashboard Cite

Rapid and cost-effective soil properties estimations are considered imperative for the monitoring and recording of agricultural soil condition for the implementation of site-specific management practices. Conventional laboratory measurements are costly and time-consuming, and, therefore, cannot be considered appropriate for large datasets. This article reviews laboratory and proximal sensing spectroscopy in the visible and near infrared (VNIR)–short wave infrared (SWIR) wavelength region for soil organic carbon and soil organic matter estimation as an alternative to analytical chemistry measurements. The aim of this work is to report the progress made in the last decade on data preprocessing, calibration approaches, and system configurations used for VNIR-SWIR spectroscopy of soil organic carbon and soil organic matter estimation. We present and compare the results of over fifty selective studies and discuss the factors that affect the accuracy of spectroscopic measurements for both laboratory and in situ applications.

show abstract

“…Malazi and Davari [16] used RF and emerging pattern algorithms to identify the complex activities of elders at home and the performance reached a high degree of accuracy with the F-measure index. Santana et al [17] quantified the quality soil parameters based on the multivariable regression of RF, making it possible to develop a fast and automatic analysis process. Anitha and Siva [18] proposed a new computer-aided brain tumor detection method using the RF classifier.…”

Section: Random Forestsmentioning

confidence: 99%

An Improved Random Forest Algorithm for Predicting Employee Turnover

Gao

Wen

Zhang

2019

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Employee turnover is considered a major problem for many organizations and enterprises. The problem is critical because it affects not only the sustainability of work but also the continuity of enterprise planning and culture. Therefore, human resource departments are paying greater attention to employee turnover seeking to improve their understanding of the underlying reasons and main factors. To address this need, this study aims to enhance the ability to forecast employee turnover and introduce a new method based on an improved random forest algorithm. The proposed weighted quadratic random forest algorithm is applied to employee turnover data with high-dimensional unbalanced characteristics. First, the random forest algorithm is used to order feature importance and reduce dimensions. Second, the selected features are used with the random forest algorithm and the F-measure values are calculated for each decision tree as weights to build the prediction model for employee turnover. In the area of employee turnover forecasting, compared with the random forest, C4.5, Logistic, BP, and other algorithms, the proposed algorithm shows significant improvement in terms of various performance indicators, specifically recall and F-measure. In the experiment using the employee dataset of a branch of a communications company in China, the key factors influencing employee turnover were identified as monthly income, overtime, age, distance from home, years at the company, and percent of salary increase. Among them, monthly income and overtime were the two most important factors. The study offers a new analytic method that can help human resource departments predict employee turnover more accurately and its experimental results provide further insights to reduce employee turnover intention.

show abstract

Visible and near infrared spectroscopy coupled to random forest to quantify some soil quality parameters

Cited by 88 publications

References 27 publications

A Regional Legacy Soil Dataset for Prediction of Sand and Clay Content with Vis-Nir-Swir, in Southern Brazil

A Regional Legacy Soil Dataset for Prediction of Sand and Clay Content with Vis-Nir-Swir, in Southern Brazil

From Laboratory to Proximal Sensing Spectroscopy for Soil Organic Carbon Estimation—A Review

An Improved Random Forest Algorithm for Predicting Employee Turnover

Contact Info

Product

Resources

About