On the selection of samples for multivariate regression analysis: application to near-infrared (NIR) calibration models for the prediction of pulp yield in Eucalyptus nitens

2008 ◽  
Vol 38 (10) ◽  
pp. 2626-2634 ◽  
Author(s):  
Christian R. Mora ◽  
Laurence R. Schimleck
Keyword(s):  
Spectral Data ◽  
Computer Aided Design ◽  
Near Infrared ◽  
Cross Validation ◽  
Sample Selection ◽  
Pulp Yield ◽  
Eucalyptus Nitens ◽  
Calibration Models ◽  
Calibration Sets ◽  
Measured Property

The effects of using reduced calibration sets on the development of near-infrared (NIR) calibration models for the prediction of kraft pulp yield in Eucalyptus nitens (Dean & Maiden) Maiden trees were explored. Three selection techniques based on NIR spectral data (CADEX (computer-aided design of experiments), DUPLEX, and SELECT algorithms) and one selection method based on a measured property (RANKING algorithm) were used for analysis and compared against a model using all data. The effect of using calibration sets of different sizes was also evaluated. All sample-selection methods resulted in models of similar performance compared with the model fitted using all samples. For calibration purposes, RANKING selection resulted in models with the lowest errors of cross-validation, followed by the DUPLEX, CADEX, and SELECT methods. In terms of validation, the SELECT and CADEX methods resulted in lower errors of prediction compared with the DUPLEX and RANKING algorithms. In general, cross-validation and prediction errors decreased as the number of calibration samples increased. These results show that it is possible to obtain adequate NIR calibration models with a reduced number of samples allowing the remaining samples to be used for model validation and that sample selection based on NIR spectral data alone is as successful as selection based on a measured property.

Prediction of neutral detergent fiber content in corn stover using near-infrared spectroscopy technique

2020 ◽  
Vol 0 (0) ◽  
Author(s):  
Xuyang Pan ◽  
Laijun Sun ◽  
Guobing Sun ◽  
Panxiang Rong ◽  
Yuncai Lu ◽  
...  
Keyword(s):  
Infrared Spectroscopy ◽  
Prediction Model ◽  
Spectral Data ◽  
Corn Stover ◽  
Near Infrared Spectroscopy ◽  
Near Infrared ◽  
Cross Validation ◽  
Fiber Content ◽  
Partial Least Square ◽  
Neutral Detergent Fiber

AbstractNeutral detergent fiber (NDF) content was the critical indicator of fiber in corn stover. This study aimed to develop a prediction model to precisely measure NDF content in corn stover using near-infrared spectroscopy (NIRS) technique. Here, spectral data ranging from 400 to 2500 nm were obtained by scanning 530 samples, and Monte Carlo Cross Validation and the pretreatment were used to preprocess the original spectra. Moreover, the interval partial least square (iPLS) was employed to extract feature wavebands to reduce data computation. The PLSR model was built using two spectral regions, and it was evaluated with the coefficient of determination (R2) and root mean square error of cross validation (RMSECV) obtaining 0.97 and 0.65%, respectively. The overall results proved that the developed prediction model coupled with spectral data analysis provides a set of theoretical foundations for NIRS techniques application on measuring fiber content in corn stover.

Evaluation of Kraft pulp yield and syringyl/guaiacyl ratio from standing trees (Eucalyptus camaldulensis, E. urophylla, Leucaena leucocephala and Casuarina junghuhniana) using portable near infrared spectroscopy

2022 ◽  
pp. 096703352110636
Author(s):  
Payyavula Ramadevi ◽  
Rathinam Kamalakannan ◽  
Ganapathy P Suraj ◽  
Deepak V Hegde ◽  
Mohan Varghese
Keyword(s):  
Leucaena Leucocephala ◽  
Kraft Pulp ◽  
Cross Validation ◽  
Nir Spectroscopy ◽  
Eucalyptus Camaldulensis ◽  
Pulp Yield ◽  
Mean Square ◽  
Calibration Models ◽  
Standing Trees ◽  
G Ratio

Measurement of pulpwood traits from a standing tree has considerable advantage when screening large populations for tree selection. It reduces time and also eliminates requirements of transport, powdering, and storing the sample. This study describes estimation of Kraft pulp yield (KPY) in Eucalyptus camaldulensis, E. urophylla, Leucaena leucocephala, and Casuarina junghuhniana by portable NIR spectroscopy of standing trees. Calibration models were developed for KPY estimation using portable NIR spectroscopy for the four species, along with a calibration model for syringyl/guaiacyl (S/G) ratio in E. camaldulensis. The calibration models for KPY showed R2 values ranging from 0.93 ( E. camaldulensis) to 0.83 ( L. leucocephala), and 0.95 for S/G ratio. The developed calibration models for E. camaldulensis and L. leucocephala were compared with laboratory NIR models, and a variation of <±2.0% was found between both methods. The models were validated by both external and cross validation which showed <2.0% RMSEP (root mean square error of prediction) and <2.0% RMECV (root mean square error of cross validation) in external and cross validations, respectively.

Estimation of whole-tree kraft pulp yield of Eucalyptus nitens using near-infrared spectra collected from increment cores

2005 ◽  
Vol 35 (12) ◽  
pp. 2797-2805 ◽  
Author(s):  
Laurence R Schimleck ◽  
Peter D Kube ◽  
Carolyn A Raymond ◽  
Anthony J Michell ◽  
Jim French
Keyword(s):  
Kraft Pulp ◽  
Near Infrared ◽  
Nir Spectroscopy ◽  
Pulp Yield ◽  
Eucalyptus Nitens ◽  
Increment Core ◽  
Traditional Assessment ◽  
Shining Gum ◽  
A Site ◽  
Whole Tree

Eucalyptus nitens (Deane and Maiden) Maiden (shining gum) is widely grown for kraft pulp production. Improving the kraft pulp yield of E. nitens increases plantation profitability but traditional assessment is slow and expensive, which hinders improvement. Near-infrared (NIR) spectroscopy provides a rapid and inexpensive method for estimating pulp yield, but studies have been limited to estimating whole-tree pulp yield using whole-tree composite samples obtained destructively. For whole-tree pulp-yield calibrations to be used non-destructively they must be applied to increment cores. In this study we used a Tasmanian E. nitens whole-tree pulp yield calibration to estimate the whole-tree pulp yields of trees from a site not included in the calibration. This was done using NIR spectra from increment cores and whole-tree composite chips. Predictions of whole-tree pulp yield based on increment cores were better than those obtained using whole-tree composite chips. The accuracy of pulp-yield predictions was greatly improved by adding a small number of prediction-set samples to the calibration sets. Calibrations for estimating whole-tree pulp yield were also obtained using NIR spectra from milled cores and whole-tree composite chips. The calibrations had similar statistics, indicating that it is possible to obtain calibrations for estimating whole-tree pulp yield based on increment-core NIR spectra.

Prediction of Chemical Composition in Sunflower Whole Plant and Silage (Helianthus Annus L.) by near Infrared Reflectance Spectroscopy

2007 ◽  
Vol 15 (3) ◽  
pp. 201-207 ◽  
Author(s):  
A. Fassio ◽  
A. Gimenez ◽  
E. Fernandez ◽  
D. Vaz Martins ◽  
D. Cozzolino
Keyword(s):  
Chemical Composition ◽  
Near Infrared ◽  
Cross Validation ◽  
Reference Method ◽  
Reflectance Spectroscopy ◽  
Least Squares Regression ◽  
Near Infrared Reflectance ◽  
Whole Plant ◽  
Calibration Models

The aim of this study was to investigate the potential use of near infrared (NIR) reflectance spectroscopy to predict chemical composition in both sunflower whole plant (WPSun) and sunflower silage (SunS). Samples of both WPSun ( n = 73) and SunS ( n = 50) were analysed by reference method and scanned in reflectance using a NIR monochromator instrument (400–2500 nm). Calibration models were developed between NIR data and reference values for dry matter (DM), crude protein (CP), ash, acid detergent fibre (ADFom), neutral detergent fibre (aNDFom), in vitro organic matter digestibility (OMD), ether extract (EE) and pH using partial least squares regression (PLS). Due to the limited number of samples full cross-validation was used to test the calibration models. The best correlations (R 2cal) and lowest standard errors in cross-validation (SECV) were obtained for DM (R 2cal > 0.82, SECV: 27.0 and 35.8 g kg−1), CP (R 2cal> 0.85, SECV: 9.9 and 10.1 g kg−1) and ash (R2cal> 0.85, SECV 11.2 and 8.2 g kg−1) in both WPSun and SunS samples, respectively. For ADFom, aNDFom and OMD the calibrations were considered to be poor (R 2cal < 0.85). In SunS samples a good correlation was found for EE (R 2cal = 0.94, SECV: 15.3 g kg−1).

Determination of potentially mineralizable nitrogen and nitrogen in particulate organic matter fractions in soil by visible and near-infrared reflectance spectroscopy

2004 ◽  
Vol 142 (3) ◽  
pp. 335-343 ◽  
Author(s):  
A. MORON ◽  
D. COZZOLINO
Keyword(s):  
Near Infrared ◽  
Cross Validation ◽  
Reflectance Spectroscopy ◽  
Infrared Reflectance ◽  
Near Infrared Reflectance Spectroscopy ◽  
Near Infrared Reflectance ◽  
Calibration Models ◽  
Potentially Mineralizable Nitrogen ◽  
Mineralizable Nitrogen ◽  
Organic Matter Fractions

Visible (VIS) and near-infrared reflectance spectroscopy (NIRS) combined with multivariate data analysis was used to predict potentially mineralizable nitrogen (PMN) and nitrogen in particulate organic matter fractions (PSOM-N). Soil samples from a long-term experiment (n=24) as well as soils under commercial management (n=160) in Uruguay (South America) were analysed. Samples were scanned in a NIRS 6500 monochromator instrument by reflectance (400–2500 nm). Modified partial least square regression (MPLS) and cross validation were used to develop the calibration models between NIRS data and reference values. NIRS calibration models gave a coefficient of determination for the calibration (R2CAL)>0·80 and the standard deviation of reference data to standard error in cross validation (RPD) ratio ranging from 2 to 5·5 for the variables evaluated. The results obtained in the study showed that NIRS could have the potential to determine PMN and PSOM-N fractions in soils under different agronomic conditions. However, the relatively limited number of samples led us to be cautious in terms of conclusions and to extend the results of this work to similar conditions.

Predicting intramuscular fat, moisture and Warner-Bratzler shear force in pork muscle using near infrared reflectance spectroscopy

2006 ◽  
Vol 82 (1) ◽  
pp. 111-116 ◽  
Author(s):  
N. Barlocco ◽  
A. Vadell ◽  
F. Ballesteros ◽  
G. Galietta ◽  
D. Cozzolino
Keyword(s):  
Shear Force ◽  
Near Infrared ◽  
Cross Validation ◽  
Nir Spectroscopy ◽  
Intramuscular Fat ◽  
Coefficient Of Determination ◽  
Infrared Reflectance ◽  
Near Infrared Reflectance ◽  
Calibration Models ◽  
Intact Muscle

AbstractPartial least-squares (PLS) models based on visible (Vis) and near infrared reflectance (NIR) spectroscopy data were explored to predict intramuscular fat (IMF), moisture and Warner Bratzler shear force (WBSF) in pork muscles (m. longissimus thoracis) using two sample presentations, namely intact and homogenized. Samples were scanned using a NIR monochromator instrument (NIRSystems 6500, 400 to 2500 nm). Due to the limited number of samples available, calibration models were developed and evaluated using full cross validation. The PLS calibration models developed using homogenized samples and raw spectra yielded a coefficient of determination in calibration (R2) and standard error of cross validation (SECV) for IMF (R2=0·87; SECV=1·8 g/kg), for moisture (R2=0·90; SECV=1·1 g/kg) and for WBSF (R2=0·38; SECV=9·0 N/cm). Intact muscle presentation gave poorer PLS calibration models for IMF and moisture (R2<0·70), however moderate good correlation was found for WBSF (R2=0·64; SECV=8·5 N/cm). Although few samples were used, the results showed the potential of Vis-NIR to predict moisture and IMF using homogenized pork muscles and WBSF in intact samples.

Evaluation of salt content of curry soup containing coconut milk by near infrared spectroscopy

2018 ◽  
Vol 26 (3) ◽  
pp. 149-158 ◽  
Author(s):  
Ekkapong Cheevitsopon ◽  
Panmanas Sirisomboon
Keyword(s):  
Infrared Spectroscopy ◽  
Root Mean Square Error ◽  
Near Infrared Spectroscopy ◽  
Near Infrared ◽  
Cross Validation ◽  
Salt Content ◽  
Coconut Milk ◽  
Mean Square ◽  
Test Set ◽  
Calibration Models

A feasibility study was performed to assess whether near infrared spectroscopy could evaluate the salt content of curry soup containing coconut milk. The soup samples were from the mixing tank, a water content adjusted tank, the ultra-high temperature pipe, and laminated containers of a food processor plant. In addition, fish sauce adjusted samples made from the same recipe but with increasing or decreasing (±30%, 60%, and 90%) sauce content were prepared. There were 113 samples in total, which were scanned using a Fourier-transform near infrared spectrometer. The prediction models for salt content were established using near infrared spectral data in conjunction with partial least squares regression. Calibration models developed using all of the samples were validated using leave-one-out cross validation and test set validation. The unadjusted sample models were validated using test set validation. The results showed that both validation methods for the calibration models using all of the samples provided similar model performance where the r2, root mean square error of calibration/root mean square error of prediction, and residual predictive deviation were 0.956, 0.065%, and 4.77 for cross validation and 0.954, 0.064%, and 4.64 for the test set, respectively. However, the salt unadjusted sample model showed better performance where the r2, RMSEP, and RPD were respectively 0.963, 0.043%, and 5.23, indicating that excellent models can be developed to determine the salt content of curry soup containing coconut milk for any applications, including quality assurance.

Estimation of grape Isabel's quality in pre-harvest based on spectral signature

2020 ◽  
Vol 1 (1) ◽  
pp. 011-013
Author(s):  
Maria Aparecida Lima ◽  
Antônio Odair Santos
Keyword(s):  
Phenolic Compounds ◽  
Spectral Data ◽  
Near Infrared ◽  
Motor Vehicle ◽  
Spectral Signature ◽  
Grape Quality ◽  
Calibration Models ◽  
Non Destructive

A non-destructive technique, that estimate the quality of the grape, was used in a vineyard, aiming to evaluate the logistics of use a portable infrared equipment [NIRs] (Brimrose Corp, USA) incorporated in a motor vehicle. The equipment estimate in the field quantifies phenolic compounds from the Isabel grape through spectral data. The spectra were taken in clusters of grapes, in the 2017 and 2018 harvests. The Near Infrared instrumentations were connected to a laptop and positioned in a micro-tractor (Gator-John Deere) to travel the vineyard. The phenolic compounds were estimates using predetermined calibration models. This methodology proved to be promising for estimating the grape quality.

Sign in / Sign up

Export Citation Format

Share Document