An efficient method of wavelength interval selection based on random frog for multivariate spectral calibration

Yun, Yong‐Huan; Li, Hongdong; Wood, Lincoln C.; Fan, Wei; Wang, Jiajun; Cao, Dong‐Sheng; Xu, Qingsong; Liang, Yi‐Zeng

doi:10.1016/j.saa.2013.03.083

Cited by 169 publications

(85 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Random frog is a reversible jump Markov Chain Monte Carlo (RJMCMC)-like algorithm based on Bayesian approaches that was originally proposed to apply into gene selection by Li . It borrows the framework of reversible jump MCMC, conducts a search in the model space through both fixeddimensional and trans-dimensional moves between different models, and then a pseudo-MCMC chain is computed and used to output a selection probability for each variable, describing importance of variables (Yun et al 2013). The larger the probability is, the more important the corresponding variable would be.…”

Section: Chemometrics Methodsmentioning

confidence: 99%

“…According to the research of Li and Yun et al (Li et al 2012;Yun et al 2013), there were five tuning parameters controlling the performance of RF. The most important two parameters were the number of iterations N and the number of variables contained in the initialized variable set Q.…”

Section: Preprocessing Of Nir Spectroscopymentioning

confidence: 99%

“…The Q was investigated by different values of 2, 10, 30, 50, and 100 by the performance of RMSECV, and the lowest RMSECV was obtained when Q was 100 in this work. The other three parameters θ (a factor controlling the variance of a normal distribution from which the number of variables is sampled to enter a candidate variable subset), ω (the role of this parameter is explained by Yun et al (Yun et al 2013)), and η (it is the upper bound of the probability for accepting a candidate variable subset whose performance is not better than initial variable subset) not having important influences on the results were set to default value 3, 0.3, and 0.1 Yun et al 2013). The average selection probability of K at Q=100 is shown in Fig.…”

Section: Preprocessing Of Nir Spectroscopymentioning

confidence: 99%

“…Furthermore, the practice (Liu et al 2014b) has demonstrated that an efficient variable selection can be greatly beneficial both to improve the predictive ability of the model and to highly reduce its complexity. Several techniques, competitive adaptive reweighted sampling (CARS) (Li et al 2009;Fan et al 2012), genetic algorithms (GA) (Leardi 2000), and random frog (RF) Yun et al 2013), applied to spectral data and devoted to variable selection have been presented and proved effective for extracting useful variables and improving the prediction ability of PLSR models. Thus, the objective of this study was to investigate the possibility of applying NIRS combined with variable selection methods for quantitative analysis of TDF and mineral elements in Coix seed.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Rapid Determination of Total Dietary Fiber and Minerals in Coix Seed by Near-Infrared Spectroscopy Technology Based on Variable Selection Methods

et al. 2014

View full text Add to dashboard Cite

The feasibility of near-infrared (NIR) spectroscopy for determining total dietary fiber (TDF) and mineral elements (mainly K, Mg, P, and S) in Coix seed was investigated. Partial least squares regression (PLSR) was applied to establish quantitative models. Norris derivative smoothing (NDS) was used as pretreatment method. A comparison of three variable selection methods, namely competitive adaptive reweighted sampling (CARS), genetic algorithms (GA), and random frog (RF), showed that CARS obtained the best performances of PLSR models with the effective wavelengths mainly concentrated on around 12,000-11,000 cm −1 and 6500-3600 cm −1 . For the quantitative determination models of TDF, K, Mg, P, and S, the optimal root mean square error of prediction (RMSEP) values were 0. 0923, 182.7224, 75.4987, 162.6993, and 36.6278; the r values were 0.95, 0.88, 0.80, 0.96, and 0.96; the residual predictive deviation (RPD) values were 2.68, 2.05, 1.70, 3.24, and 3.04, respectively. It is concluded that the NIR spectral technique has a potential to determine TDF, K, Mg, P, and S in Coix seed.

show abstract

Section: Chemometrics Methodsmentioning

confidence: 99%

Section: Preprocessing Of Nir Spectroscopymentioning

confidence: 99%

Section: Preprocessing Of Nir Spectroscopymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Rapid Determination of Total Dietary Fiber and Minerals in Coix Seed by Near-Infrared Spectroscopy Technology Based on Variable Selection Methods

et al. 2014

View full text Add to dashboard Cite

show abstract

“…RF is a novel and efficient technique for variable selection, which borrows the framework of reversible jump Markov Chain Monte Carlo (RJMCMC) methods (Li et al 2012;Yun et al 2013). The details of RF algorithm can be found in literature (Li et al 2012;Yun et al 2013). RF works in four steps:…”

Section: Variable Selectionmentioning

confidence: 99%

Detection of Knot Defects on Coniferous Wood Surface Using Near Infrared Spectroscopy and Chemometrics

et al. 2016

View full text Add to dashboard Cite

Lumber pieces usually contain defects such as knots, which strongly affect the strength and stiffness. To develop a model for rapid, accurate grading of lumbers based on knots, Douglas fir, spruce-pine-fir (SPF), Chinese hemlock, and Dragon spruce were used. The experiments explored the effects of modelling methods and spectral preprocess methods for knot detection, and investigated the feasibility of using a model built within one species to discriminate the samples from other species, using a novel variable selection method-random frog (RF)-to select effective wavelengths. The results showed that least squares-support vector machines coupled with first derivative preprocessed spectra achieved best performance for both single and mixed models. Models built within Dragon spruce could be used to classify knot samples from SPF and Chinese hemlock but not Douglas fir, and vice versa. Eight effective wavelengths (1314 nm, 1358 nm, 1409 nm, 1340 nm, 1260 nm, 1586 nm, 1288 nm, and 1402 nm) were selected by RF to build effective wavelengths based models. The sensitivity, specificity, and accuracy in the validation set were 98.49%, 93.42%, and 96.30%, respectively. Good results could be obtained when using data at just eight wavelengths, as an alternative to evaluating the whole spectrum.

show abstract

Applications of Handheld Near‐Infrared Spectrometers

Yan

Siesler

2021

Portable Spectroscopy and Spectrometry

View full text Add to dashboard Cite

An efficient method of wavelength interval selection based on random frog for multivariate spectral calibration

Cited by 169 publications

References 43 publications

Rapid Determination of Total Dietary Fiber and Minerals in Coix Seed by Near-Infrared Spectroscopy Technology Based on Variable Selection Methods

Rapid Determination of Total Dietary Fiber and Minerals in Coix Seed by Near-Infrared Spectroscopy Technology Based on Variable Selection Methods

Detection of Knot Defects on Coniferous Wood Surface Using Near Infrared Spectroscopy and Chemometrics

Applications of Handheld Near‐Infrared Spectrometers

Contact Info

Product

Resources

About