Performing sequential forward selection and variational autoencoder techniques in soil classification based on laser-induced breakdown spectroscopy

Abstract

The feasibility and accuracy of several combination classification models, i.e., quadratic discriminant analysis (QDA), random forest (RF), Bernoulli naïve Bayes (BNB), and support vector machine (SVM) classification models combined with...

Application of laser-induced breakdown spectroscopy with a generalized regression neural network and LASSO-type methods for estimation of arsenic and chromium in soil

“…The data analysis was performed in Python 3.9.10 soware (Python Soware Foundation, USA). The performance of the proposed models was examined by using the coefficient of determination (R 2 ), root mean square error (RMSE), relative standard deviation (RSD), and limit of detection (LoD), which are described in eqn ( 8)- (12). R 2 provides a reasonable estimate of the univariate analysis and multivariate analysis.…”

“…The composition and concentration of the elements can be derived from the location and signal intensity of the analyzed lines. LIBS technology has a wide range of applications in the analysis of the substance composition and has been widely used in agricultural soil, 12 steel industry, 13 nuclear industry, 14 food safety, 15 biomedical diagnostics, 16 environmental monitoring, 17 and many other fields. For LIBS applications, quantitative analysis methods are particularly critical.…”

Fast recognition model of Fusarium in agaric based on hyperspectral imaging

“…In this subsection, the time and species models of agaric infected with Fusarium were discussed to obtain the optimal time classification models for agaric infected with two species of Fusarium, with 2ND treatment for 100% accuracy. SFS is an algorithm used for dimensionality reduction and is suitable for filtering feature data by comparing selected initial feature values with other feature values to finally identify the feature variables that meet the requirements (Edward & Weidong, 2021;Sahameh et al, 2021). The ELM is a classifier with fast learning capability and good adaptation to the model, related to the number of hidden neural nodes, of which the parameters set in this study are determined by interval cyclic selection (Jixiang et al, 2023;Satapathy et al, 2019).…”

Quantitative analysis of Cr in soil using laser induced breakdown spectroscopy combined with relevance vector machine method

“…Traditionally, basic calibration method and internal calibration method [10,11] are widely used in the quantitative analysis of laser induced breakdown spectroscopy, but the analytical accuracy and repeatability is not very satisfactory. Therefore some methods have been combined with LIBS in order to improve the quantitative analytical results of LIBS, such as Partial Least Squares method(PLS) [12][13][14][15], Artificial Neural Network (ANN) [16], Random Forest (RF) [17,18], Support Vector Machine(SVN) [19][20][21], Least Squares Support Vector Machine(LSSVM) [22,23], Relevance Vector Machine(RVM) [24] and so on. *wdzhou@zjnu.cn,; Phone 86 579 82298913…”