Comparing spectral models to predict manganese content in Rosa spp. leaves using VIS-NIR data

Principal component regression (PCR) and partial least squares (PLS) chemometric methods were applied to the simultaneous quantitative analysis of levamisole (LVM) and triclabendazole (TCB) in tablets without using a preliminary separation, even in presence of the overlapping spectra of the above compounds. For both PCR and PLS, a concentration set containing 25 different mixtures of LVM and TCB in the linear concentration range was symmetrically prepared and then the absorbance values of the concentration set were measured at the wavelength set with Dl=0.1 nm in the spectral region of 225-322.3 nm. PCR and PLS calibrations were obtained by applying the PCR and PLS algorithms to the concentration set data (y-block) and their corresponding absorbance data (x-block). The validity of PCR and PLS chemometric methods was performed by using the independent synthetic mixtures and the standard addition technique. Then, these analytical methods were applied to the commercial tablets and a good agreement was obtained between experimental results provided by the application of the PCR and PLS to the synthetic and real samples.

The quality of cocoa beans can be determined in various ways, and two of them are: (i) manual observation via splitting cocoa beans in order to determine the degree of fermentation and observe the defect; (ii) chemical analyses for determining the fat and moisture content; with the latter is known as a time-consuming process. The NIRS instrument is a kind of a non-destructive measurement that can predict rapidly the quality of cocoa beans. This study aims to simulate a mathematical model for the prediction of moisture- and fat-content using a NIRS instrument. These results were subsequently analyzed with two types of multivariate regression analysis: Principal Component Regression (PCR) and Partial least Square Regression (PLSR) and the results shown based on two methods were then compared. The PCR method delivered a higher determination coefficient in moisture analysis compared to PLSR. On the other hand, a greater determination coefficient was delivered by PLSR in terms of fat analysis compared to the value obtained via PCR. The root means square error of the PCR method was lower than that of PLSR. It can be concluded that the PLSR method is more suitable for fat content prediction.

To ensure that the plasma reactor tool can simulate the Plasma Activated Water (PAW) liquid accurately. To ensure the quality of the PAW liquid according to the plasma reactor design, it is necessary to create a calibration model that can ensure that the model matches the observation data and improve the predictive ability of the model. The purpose of this article is to build a calibration model based on the quality of PAW liquid produced from a plasma reactor. This study used the Principal Component Regression (PCR) method and the Partial Least Square Regression (PLS-R) method. The advantage of the PCR method is that it reduces data based on correlation values. While the PLS-R method reduces data based on the most relevant factors in interpreting the data. Based on the experiments conducted, it was concluded that to build a calibration model based on plasma reactor data and PAW liquid data, the PCR method is better than the PLS-R method. This is shown based on the RMSEP and R2 values in the PCR method of 0.09625571 and 93.04699% while in the PLS-R method of 0.09873341 and 92.84436%. For the R2 value in the PCR method is greater which indicates that the data variant value is more acceptable to the calibration model than in the PLS-R method, then the RMSEP value in the PCR method is smaller which indicates that the statistical error value is more acceptable than PLS-R.

Weather index insurance is a relatively new alternative to traditional agricultural insurance such as individual yield-based crop insurance. It is still mostly at the experimental stage, rather than in widespread use like traditional crop insurance. A major challenge for weather insurance is basis risk, where the loss estimated by the index differs from the actual loss, and this is generally believed to be the main limitation in the use of weather index insurance for crops. Variable basis risk is an important type of basis risk that occurs when there are incorrect variables or missing variables for the design of the weather index. In agriculture, there is a relatively small sample size of yields, and therefore, as the number of considered weather variables increases, the problems of limited degrees of freedom for predictive models must be overcome. The objective of this article is to demonstrate two possible approaches that could be used to construct a multivariable weather index to reduce variable basis risk. Forage insurance is used as an example, and a main focus of the research is on reducing the dimensionality of the predictive model and resolving the problem of multicollinearity among weather variables. The research uses daily weather information and county-level forage yield data from Ontario, Canada. Two multivariable indices are developed based on principal component regression (PCR) and partial least squares regression (PLSR) methods, and they are evaluated against a single-variable benchmark index based on cumulative precipitation (SVCP) using several basis risk metrics. The results show that both the PCR and PLSR models are superior compared to the single-variable weather index based on cumulative rainfall (SVCR) index and can be used to achieve the objective of reducing the dimensionality of the weather variable matrix and addressing the issue of multicollinearity. Though the PLSR indices perform better than the PCR and SVCR indices in terms of average value of basis risk (),(rom) the PCR method produces a smaller percentage of mismatch, suggesting that the PCR method may be superior in correctly detecting when the insurance payment should be triggered. The methods demonstrated in this article will assist in the development of weather index–based crop insurance.

Meng Ruifeng, Zhong Jianjun, Zhang Lifen, Ye Xingqian, Liu Donghong* (School of Biosystems Engineering and Food Science, Zhejiang University, Hangzhou 310058, China) Abstract: This work demonstrated the use of multivariate statistical techniques called principal component (PC) and partial least squares (PLS) to extract the acoustic features of citrus pectin water solution. The concentration of citrus pectin water solution was predicted by PC and PLS regression method using the spectra of ultrasound pulse echoes travelling through mixtures. The values of root mean square error of validation (RMSEV) were 0.0675 g/100 g and 0.0662 g/100 g for PC and PLS regression model, respectively. Since the response variable was taken into account, PLS regression model was more accurate than PC regression model. Also, a method for temperature compensation was proposed to correct the impact of temperature variation on analyzed data. The proposed methods for pectin concentration measurement are easily adaptable to similar applications using existing hardware. Keywords: Partial Least Square Regression, Principal Component Regression, concentration measurement, acoustic velocity DOI: 10.3965/j.ijabe.20120502.010 Citation: Meng R F, Zhong J J, Zhang L F, Ye X Q, Liu D H. Ultrasonic concentration measurement of citrus pectin aqueous solutions using PC and PLS regression. Int J Agric & Biol Eng, 2012; 5(2): 76

The objective of this study was to compare the predictions of the protein contents and hardness values by partial least squares regression (PLSR) and principal components regression (PCR) models for bulk samples of Canadian wheat, which were obtained from different locations and crop years. Wheat samples of Canada Western Red Spring (CWRS), Canada Western Hard White Spring (CWHWS), Canada Western Soft White Spring (CWSWS), and Canada Prairie Spring Red (CPSR) classes were obtained from nearby agricultural farms in the main wheat growing locations in the Provinces of Alberta, Saskatchewan, and Manitoba from 2007, 2008, and 2009 crop years. Wheat samples were conditioned to moisture levels of 13, 16, and 19 % (wet basis) and pooled together for developing the regression models. A database of the near-infrared (NIR) hyperspectral image cubes of bulk samples of wheat classes was created in the wavelength region of 960–1,700 nm with 10 nm intervals. Reference protein contents and hardness values were determined using the Dumatherm method and single kernel characterization system (SKCS), respectively. A tenfold cross-validation was used for the ten-factor partial least squares regression (PLSR) and principal components regression (PCR) models for prediction purposes. Prediction performances of regression models were assessed by calculating the estimated mean square errors of prediction (MSEP), standard error of cross-validation (SECV), and correlation coefficient (r). Using the full data set in the protein prediction study, the ten-component PLSR model gave 1.76, 1.33, and 0.68 for the estimated MSEP, SECV, and r, respectively, which were better than the results for the ten-component PCR model (2.02, 1.42, and 0.62, respectively). For the hardness prediction, the estimated MSEP, SECV, and r values were 147.7, 12.15, and 0.82, respectively, for the ten-component PLSR model using the full data set. The PLSR models prediction performances outperformed the PCR models for predicting protein contents and hardness of wheat.

Objective: The objective of this study was to develop a UV spectroscopy method in combination with multivariate analysis for determining vitexin in binahong (Anredera cordifolia (Ten.) Steenis) leaves extract.
 Methods: The partial least square (PLS) regression and the principal component regression (PCR) was performed in this study to evaluate several statistical performances such as coefficient of determination (R2), root mean square error of calibration (RMSEC), root mean square error of cross-validation (RMSECV), root mean square error of prediction (RMSEP) and relative error of prediction (REP). Cross-validation in this study was performed using leave one out technique.
 Results: The R2 values of calibration data sets resulted from PLS ​​and PCR method were 0.9675 and 0.9648, respectively. The low values of RMSEC and RMSECV both for PLS ​​and PCR method indicated the minimum error of the calibration models. The R2 values of validation data sets resulted from PLS ​​and PCR method were 0.9778 and 0.9820, respectively. The low values of RMSEP both for PLS ​​and PCR method indicated the minimum error of prediction generated from the calibration data sets. Multivariate calibration techniques were applied to determine the content of vitexin in binahong leaves extract. Predicted values from the multivariate calibration models were compared to the actual values determined from a validated HPLC method. It was found that PLS models resulted in the lowest REP values compared to the PCR models.
 Conclusion: The chemometrics technique can be applied as an alternative method for determining vitexin levels in the ethanol solution of binahong leaves extract.

This study aims to assess Near infrared (NIR) spectroscopy to predict vanillin and moisture content non-destructively. Twenty-four samples of vanilla pods (@55 grams/sample) from Sumatra and Sulawesi Island were tested. The reflectance spectra of vanilla pod were measured using a NIRFlex N-500 spectrometer at the wavelength of 1000-2500 nm. After that, the vanillin and moisture content were measured using the reference method. The reflectance spectra were transformed to absorbance spectra and several pre-treatment methods of NIR spectra were applied, and the results were calibrated with chemical data using principal component regression (PCR) and partial least square (PLS) methods. The best estimation for vanillin content using absorbance spectra with OSC pre-treatment at 3 PLS factors with the accuracy parameters of r = 0.91, SEC = 1.31%, SEP = 1.34%, CV = 45.60%, RPD = 2.30 with a consistency of 97.96%. The best prediction of vanillin content using PCR method was obtained with SNV pre-treated spectra at 6 PCR factors with the accuracy parameters of r = 0.89, SEC = 1.47%, SEP = 1.49%, CV = 43.21%, RPD = 2.09 and 101.40% consistency. The best estimation for water content using the PLS method is using the SNV pre-treatment spectra with a PLS factor of 5 with the accuracy parameters of r = 0.99, SEC = 2.11%, SEP = 2.13%, CV = 6.12%, RPD = 7.98 with consistency = 98.86%. While using the PCR method, the best estimation for moisture content was also obtained by SNV pre-treated spectra at 4 PCR factors with the accuracy values of r = 0.99, SEC = 2.33%, SEP = 2.25%, CV = 6.25%, RPD = 7.58 and consistency = 103.90%. The PLS calibration provide a better accuracy than the PCR. The NIR spectroscopy associated with the selected pre-treatments and factor numbers of PLS and PCR can be used for determination of chemical content of vanilla pods.

Infrared (IR) spectroscopy can serve as a rapid method for the quantitative analysis of borophosphosilicate glass (BPSG) films on Si wafers for the microelectronics industry. The advantages of using statistically designed calibration sets are emphasized. Classical least-squares (CLS), partial least-squares (PLS), and principal component regression (PCR) methods are all found to provide improved precision over traditional peak-height measurements. The quantitative results from spectral measurements taken in transmission mode at both 0/sup 0/ and 60/sup 0/ incident angles were also compared. PLS and PCR methods yielded results that were comparable within the sampling error, and each exhibited a better analysis precision than that obtained from the CLS analysis. Both PLS and PCR methods yielded the best results when applied to the original 60/sup 0/ incident angle data, which was not corrected for film thickness. PLS and PCR analyses each gave a standard error of prediction (SEP) for boron of approx. = 0.1 wt% and approx. = 0.2 wt % for phosphorus for a set of 44 calibration samples which spanned a range of concentrations from 1 to 5 wt % B and 2 to 6 wt % P. The PLS and PCR methods applied to the IR spectra were also capablemore » of monitoring film thickness with a SEP of 14 nm for films that varied in thickness from 430 to 1000 nm. The importance of using these full-spectrum multivariate methods for outlier sample detection is presented, and the ability to extract qualitative spectral information from the CLS and PLS calibrations is demonstrated.« less

In this study, a model for prediction of lignocellulose components of agricultural residues has been developed with Fourier Transformed Near Infrared (FT-NIR) spectroscopy data. Two calibration techniques (Principal Component Regression (PCR) and Partial Least Square Regression (PLSR)) were assessed for prediction of lignin, holocellulose, α-cellulose, pentosan and ash, and found the PLSR better for lignin, holocellulose and α-cellulose. The PCR also produced better results for quantification of pentosan and ash. Spectral range (7000–5000 cm–1) showed more informative than other parts of the spectral data. The PLSR showed maximum value of R2 (R2 = 0.91%) for prediction of holocellulose. For the prediction of pentosan, the PCR was better (R2 = 0.68%). The PCR also showed better results (R2 = 86%) for quantification of ash. To determine amount of lignin, the PLSR was the best (R2 = 0.83%) when the spectral data were de-trained and smoothed with Savitzky-Golay (S-G) filtering simultaneously. For prediction of α-cellulose, the PLSR was the best model (R2 = 0.94%) when the data were pretreated with mean normalization. Considering the best alternatives inNear Infrared (NIR) data preprocessing and calibration techniques, methods for quantification of lignocellulose components of agricultural residues have been developed which is rapid, cost effective, and less chemical intensive and easily usable in pulp and paper industries and pulp testing laboratories.

The aim of this work was to compare principal component regression (PCR) and partial least squares (PLS) regression methods while estimating the piperine content in black pepper using Raman spectroscopy. The calibration and prediction models of the regression analysis on Raman spectra were developed using PCR and PLS algorithm. The efficiency of the developed models was evaluated by means of root mean square error of calibration (RMSEC), root mean square error of prediction (RMSEP), and correlation coefficient (R2). For PCR algorithm, these parameters were obtained as 0.1, 0.1, and 0.97, respectively; and for PLS regression, the parameters were found as 0.05, 0.08, and 0.99, respectively. The results revealed that Raman spectroscopy with PCR and PLS algorithm could be used for determining the concentration of piperine in black pepper with an accuracy of 92.35% and 94.74% respectively.

Accelerated lattice constant prediction of perovskite materials (ABX3, A2BB′O6) using partial least squares and principal component regression methods

Rapid prediction of multiple wine quality parameters using infrared spectroscopy coupling with chemometric methods

Partial least-squares (PLS) methods for quantitative spectral analyses are compared with classical least-squares (CLS) and principal component regression (PCR) methods by using simulated data and infrared spectra from bulk seven-component, silicate-based glasses. Analyses of the simulated data sets show the effect of data pretreatment, base-line variations, calibration design, and constrained mixtures on PLS and PCR prediction errors and model complexity. Analyses of the simulated data sets also illustrate some qualitative differences between PSL and PCR. PLS and PCR predicted concentration errors from the simulated data sets and a set of the Fourier transform infrared spectra of silicate-based glasses (S-glass) show that prediction errors are not statistically different between these two methods for these individual data sets with limited numbers of samples. However, PLS and PCR are both superior to CLS methods in the case of the analysis of S-glass where only one analyte is known in the calibration samples and the components of unknown concentration overlap all the spectral features of the analyte components. CLS analysis precision significantly improves when the three known analyte concentrations (B/sub 2/O/sub 3/, P/sub 2/O/sub 5/, and OH) are used in the calibration. In this latter case, PLS and PCR concentration predictions are unchanged, andmore » although they each still yield a lower standard error of prediction than the CLS method, there is no longer strong statistical evidence that these differences between PLS or PCR and CLS are outside experimental error for the B/sub 2/O/sub 3/ component. The ability of CLS and PLS methods to provide chemically useful estimates of the pure-component spectra is also demonstrated.« less

Partial Least Squares (PLS) and Principal Component Regression (PCR) methods were applied to the simultaneous determination of a mixture of twelve pesticides by high performance liquid chromatography (HPLC). Calibration models at two different wavelengths were developed to resolve mixtures of the pesticides with overlapping chromatographic peaks. The first model carried out at 205 nm, as first detector compromise wavelength, yielded satisfactory sensitivity and selectivity for estimation of the concentration of iprodione, procymidone, triadimefon, and vinclozolin. The other model at 250 nm, as second detector compromise wavelength, was used for estimation of chlorothalonil, clorfenvinphos, fenamiphos, parathion-methyl, parathion-ethyl, and triazophos. However, two pesticides of the mixture, malathion and tebuconazole, showed bad prediction ability and were not determined, perhaps owing to their low signal relative to the other compounds. Both calibration models were evaluated by predicting the concentration of independent test set samples, and were successfully applied to the determination of these pesticides in groundwater samples. In all cases the PLS calibration method showed superior quantitative prediction ability than the PCR method.