Interval Partial Least Squares Research Articles

Quantitative Raman spectroscopy for the Ioncell\xae process: Part 2\u2014quantification of ionic liquid degradation products and improvement of prediction performance through interval PLS

One of the main issues associated with ionic liquids (ILs) is their recyclability. Viable recycling concepts can only be developed if one knows what is in the IL mixtures and solutions. In our previous work, we showed that it is possible to quantify water and 1.5-diazabicyclo[4.3.0]non-5-enium acetate [DBNH][OAc] IL components in liquid mixtures using Raman spectroscopy. In this regard, we considered Raman spectroscopy as a promising analytical method for the inline monitoring and control of the Ioncell® process. In the present work, we push the limits of this analytical method further by extending it to more complex and realistic liquid mixtures including the hydrolysis product 1-(3-aminopropyl)-2-pyrrolidone (APP) that can be formed upon the reaction of 5-diazabicyclo[4.3.0]non-5-ene (DBN) with water. Quantifying APP is important in order to measure the extent of the hydrolysis reaction and apply the right corrective measures to reverse the reaction and to maintain the process within the optimal working conditions. The simultaneous quantification of the four components (Acetic acid, DBN, APP and H2O) in typical Ioncell® liquid streams is investigated using Raman spectroscopy. The sensitivity of the Raman method in quantifying APP is also highlighted in comparison with refractometry, which is widely applied to measure IL concentration in aqueous mixtures. Finally, we propose simple modifications on the multivariate partial least square regression model based on a variable selection algorithm to enhance the accuracy of the predicted calibration values.

Exploring multivariate linear regression methods for the prediction of total phenolic content in standard American lager beers using synchronous fluorescence spectroscopy fused data

This work aims to explore chemometric methods for determining the total phenolic content in Brazilian standard American lager beers using synchronous fluorescence spectroscopy fused data. Two hundred and sixteen (216) beer samples were diluted at a ratio of 1:5 (v/v), sonicated for 15 ​min in a water bath, and buffered using an acetic acid/acetate buffer of pH 4.5. The synchronous fluorescence spectra were obtained from 200 to 600 ​nm with a resolution of 1 ​nm. However, different offsets were tested for synchronous fluorescence (Δλ ​= ​20, 50, and 80). All measurements were carried out in triplicate, and the Folin-Ciocalteu method was used as the reference method for the total phenolic content. All chemometric methods were explored using R software and its graphic interface, RStudio. First, the spectral data with different offsets were normalized and fused. After this, the fused data were smoothed with the Savitzky-Golay algorithm with a window of length 11 and a quadratic polynomial, followed by data mean centering. Robust principal component analysis (ROBPCA) was used to identify anomalous data. The Kennard-Stone algorithm was used to split the dataset into 75% for the training set and 25% for the test set. The prediction models used were partial least squares (PLS), orthogonal partial least squares (O-PLS), and elastic net regressions with cross-validation (CV). Ensemble PLS and ensemble sparse PLS with bootstrap resampling were also used. Other PLS regression methods based on the selection of variables were tested, such as significance multivariate regression (sMC – PLS), uninformative variable elimination (UVE – PLS), and interval PLS (iPLS) global and optimized models with cross-validation. All the regression models were built with the training set. The validation was carried out with the application of the regression models to the test set. Within this context, two multivariate regression were highlighted, the CV elastic net and iPLS optimized models, with the root mean square error of prediction (RMSEP) equaling 1.0 and 1.1 ​mg GAE L−1, respectively.

Rapid quantification of constituents in tobacco by NIR fiber‐optic probe

AbstractThis study aims to develop, evaluate, and optimize the application of near‐infrared spectroscopy (NIR) autosampler device equipped with a fiber‐optic probe associated with chemometric methods for the fast quantification of routine chemical constituents such as total alkaloids, reducing sugars, nitrate, and ammonia in tobacco. For this purpose, an NIR autosampler device with capacity for 40 samples was used. The spectra were collected in the range of 950 to 2,500 nm, with three spins on the sample, a 3.0‐cm2 sample scanning area, and 2.0 ± 0.2 mm distance between the fiber‐optic probe and the sample. Four partial least‐squares (PLS) models were developed, and different preprocessing methods were investigated. The predicted results were compared with those obtained from the reference method (continuous‐flow analysis), and root mean square error of prediction values of 0.31%, 1.27%, 0.47%, and 0.026% were obtained for total alkaloids, reducing sugars, nitrate, and ammonia, respectively. The proposed method performed well for the analysis of total alkaloids and reducing sugars with an appropriate goodness‐of‐fit and fair precision. In conclusion, considering the performance of the regression models and the associated environmental and economic advantages, the application of NIR spectrometer autosampler device equipped with a fiber‐optic probe associated with a PLS and synergy interval PLS algorithms cannot replace the reference method, but it is a promising tool for tobacco monitoring.

Improved prediction of ‘Kent’ mango firmness during ripening by near-infrared spectroscopy supported by interval partial least square regression

Mangoes (Mangifera indica L.) are tropical fruits, which are sourced worldwide to supply the consumer market in Europe. Often mangoes are transported over sea in refrigerated containers at 8–10 °C and in some cases under controlled atmosphere conditions. At arrival in European countries, a vast amount of fruit is ripened under specified conditions to deliver 'Ready to Eat' fruit to consumers. The latter is a challenge due to great variability in fruit maturity stage at arrival. There are currently no good methodologies to rapidly and nondestructively monitor and control the ripening process of mangoes. A major indicator of mango ripeness is fruit firmness. In the present study, a portable visible near-infrared (400–1130 nm) (VNIR) spectrometer was used to predict the firmness of individual mango undergoing ripening. Ripening of ‘Kent’ mango was for 10 days monitored at 20 °C and relative humidity (RH) of 85%. Every other day fruit firmness (measured with AWETA acoustic firmness analyzer) and NIR spectrum were determined on 2 opposite sides of the fruit. Interval partial-least square (iPLSR) regression was used for identifying the important wavelengths responsible for predicting firmness in mangoes. Results showed a change in the VNIR spectra with the change in firmness of mangoes. The model based on selected wavelengths performed significantly better compared to PLSR without pre-selecting wavelengths. iPLSR based regression provided a correlation of calibration and prediction as R2c = 0.75 and R2p = 0.75, and root means squared error of calibration and prediction as 6.02 Hz2g2/3 and 5.92 Hz2g2/3 respectively. The iPLSR model outperformed the standard PLSR model by over 12% in R2p and 14% reduction in prediction error. The predictions by the model provided an evolution of the firmness during the complete ripening experiment. Non-destructive access to mango firmness during ripening can assist in optimizing the process to better meet the market demand.

An Effective Prediction Approach for Moisture Content of Tea Leaves Based on Discrete Wavelet Transforms and Bootstrap Soft Shrinkage Algorithm

The traditional method used to determine the moisture content of tea leaves is time consuming and destructive. To address this problem, an effective and non-destructive prediction method based on near-infrared spectroscopy (NIRS) is proposed in this paper. This new method combines discrete wavelet transforms (DWT) with the bootstrap soft shrinkage algorithm (BOSS). To eliminate uninformative or interfering variables, DWT is applied to remove the noise in the spectral data by decomposing the origin spectrum into six layers. BOSS is used to select informative variables by reducing the dimensions of the sub-layers’ reconstruction spectrum. After selecting the effective variables using DWT and BOSS, a prediction model based on partial least squares (PLS) is built. To validate effectiveness and stability of the prediction model, full-spectrum PLS, genetic algorithm PLS (GA-PLS), and interval PLS (iPLS) were compared with the proposed method. The experiment results illustrate that the proposed prediction model outperforms the other classical models considered in this study and shows promise for the prediction of the moisture content in Yinghong No. 9 tea leaves.

Chemometric-assisted UV spectrophotometric method for determination of N, N- diethyl-3-methylbenzamide in insect repellents.

In recent years outbreaks of vector-borne diseases have caused great concern to the population, especially those diseases transmitted by mosquitoes. Repellents appear as an affordable alternative for prevention, making it increasingly important to control the quality of these products, since the content of the active ingredients are directly related to the efficiency and the protection time provided by the repellent. This paper proposes an analytical method for determining the DEET (N, N- Diethyl-3-methylbenzamide) content in insect repellents in lotion using UV spectroscopy. For this propose five different strategies of regression were evaluated: (a) Partial Least Squares (PLS) using full-spectrum; (b) interval PLS (iPLS); Multiple Linear Regression (MLR) with variable selection by the (c) Genetic Algorithm (MLR/GA), (d) Successive Projections Algorithm (MLR/SPA) and the (e) Stepwise (MLR/SW). Appropriate predictions were obtained with RMSEP values between 0.88 and 0.93%w w-1. No systematic error was observed and no significant differences were found between the predicted and reference values, according to a paired t-test at 95% confidence level. The results demonstrated the potential of UV spectroscopy associated to multivariate calibration to determine DEET content in repellents as a fast, simple strategy and with a suitable correlation between the values estimated by the model and the reference values.

Exploring relevant wavelength regions for estimating soil total carbon contents of rice fields in Madagascar from Vis-NIR spectra with sequential application of backward interval PLS

ABSTRACT Laboratory visible and near-infrared (Vis-NIR) spectroscopy with partial least squares (PLS) regression can be used to determine the soil carbon (C) content, and the waveband selection procedures can refine the predictive ability. However, individually selected wavebands are not always the same depending on the location, scale, and approach. To simplify the variable selection issue, some methods for selecting wavelength regions instead of individual wavebands have been proposed. In this study, we explore relevant wavelength regions for predicting the total carbon (TC) content of lowland and upland soils in Madagascar from Vis-NIR spectroscopy using a dynamic version of backward interval PLS (biPLS) regression. The predictive ability of dynamic biPLS was compared with that of standard full-spectrum PLS (FS-PLS) using the cross-validated coefficient of determination (R 2), root mean squared error (RMSE), and ratio of performance to interquartile distance (RPIQ). The biPLS models using reflectance (R 2 = 0.877, RMSE = 0.690) and first derivative reflectance (FDR) (R 2 = 0.940, RMSE = 0.494) data sets showed better predictive accuracy than the FS-PLS models using reflectance (R 2 = 0.826, RMSE = 0.809) and FDR (R 2 = 0.933, RMSE = 0.518) data sets, the spectral efficiency was improved. By using biPLS to predict soil TC, the model was simplified by using only four selected wavelength regions in the reflectance (400–490, 1402–1440, 1846–1980 and 2151–2283 nm) and FDR (652–687, 1322–1443, 1856–1985, and 2290–2400 nm) data sets, which yielded reliable (RPIQ > 2.5) predictions.

FTIR and UV spectra for the prediction of triterpene acids in Macrohyporia cocos

The aim of this study was to exploit the potential of rapid and simple spectroscopic techniques combined with partial least squares (PLS) regression using for the content prediction of five triterpene acids in Macrohyporia cocos. Ultraviolet of 80% ethanol extract (UV1), ultraviolet of ethyl acetate extract (UV2), Fourier transform mid-infrared (FT-MIR) and Fourier transform near-infrared (FT-NIR) spectra of samples were collected. The data processing methods of spectral pretreatment, feature extraction and data fusion were adopted to improve prediction performance. The results showed that: (1) UV1, UV2 and FT-MIR did work for content prediction except FT-NIR. (2) Variable importance for the projection (VIP) outperformed interval PLS in feature extraction for the prediction of triterpene acids of M. cocos. Data fusion had no distinct advantage compared with single set analysis (3) Both UV1 and FT-MIR could achieve the prediction of poricoic acid A, and UV1 outperformed FT-MIR. (4) The UV1 subjected by second-order derivative pretreatment and VIP extraction was satisfactory for predicting poricoic acid A and dehydrotumulosic acid. The UV2 preprocessed by first-order derivative could achieve approximate prediction of dehydrotrametenolic acid. This study demonstrated that ultraviolet and mid-infrared spectral data were promising techniques for the rapid prediction of poricoic acid A (UV1 and FT-MIR), dehydrotumulosic acid (UV1) and dehydrotrametenolic acid (UV2) in M. cocos, and ultraviolet spectroscopy was superior to infrared one.

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

Infrared spectroscopy (IRs) coupling with chemometric methods were used to predict principal quality parameters in wine. A new strategy of variable (wavelength) selection named as Fisher Discriminant-Variable Selection (FD-VS) model was constructed. Characteristic variables were selected from Infrared spectra based on the absolute values of eigenvector obtained by Fisher Discriminant Function. The FD-VS method was combined with quantitative models including Principal Component Regression (PCR), Partial Least Squares (PLS) and Least Squares Support Vector Regression (LSSVR), which were utilized for prediction of multiple principal quality parameters of red wine. It is shown that FD-VS method obviously improves the performances of PCR, PLS and LSSVR models. Then four variable selection methods based on PLS regression including Competitive Adaptive Reweighted Sampling (CARS)-PLS, Uninformative Variable Elimination (UVE)-PLS, Interval Partial Least Squares (IPLS) and Moving Windows Partial Least Squares (MWPLS) were also compared. The results also show good performance of FD-VS-LSSVR in terms of prediction accuracy or robustness. Therefore, the FD-VS method provides an effective and credible variable selection way for IR spectrum to predict quality parameters of wine.

Rapid detection of saffron (Crocus sativus L.) Adulterated with lotus stamens and corn stigmas by near-infrared spectroscopy and chemometrics

Counterfeiting and adulteration have seriously affects the products quality. Saffron is the world’s most expensive medicinal plant, an excellent dye, and a high-grade spice. To reduce costs, it is often adulterated with dyed stamens of Nelumbo nucifera Gaertn. (lotus stamens) and stigmas of Zea mays L. (corn stigmas), which have a similar appearance to saffron. Therefore, the aim of this study was to identify and quantify these two adulterants in saffron using near-infrared (NIR) spectroscopy. A partial least squares discriminant analysis (PLS-DA) model was established to determine the authenticity of the saffron. A partial least squares (PLS) regression model was used to predict the adulteration level. The synergistic interval PLS (SI-PLS), competitive adaptive reweighted sampling (CARS), and Monte Carlo uninformative variable elimination variable (MC-UVE) selection methods were compared for correcting the regression models. For saffron adulterated with lotus stamens, the SI-PLS model effect was better than those of the other models. In the quantitative analysis of saffron adulterated with corn stigmas, the CARS-PLS model gave the best stability and predictability. The results showed that the variable selection methods improved the accuracy of the model. The NIR spectroscopy combined with chemometrics could effectively and rapidly identify and quantify adulteration of saffron with lotus stamens and corn stigmas without damaging the samples.

Simultaneous quantification of active constituents and antioxidant capability of green tea using NIR spectroscopy coupled with swarm intelligence algorithm

A simple, rapid and low-cost analytical method was employed for simultaneous determination of bioactive constituents and antioxidant capability of green tea. The strategy was based on swarm intelligence algorithms with partial least squares (PLS) such as simulated annealing PLS (SA-PLS), ant colony optimization PLS (ACO-PLS), genetic algorithm PLS (GA-PLS), and synergy interval PLS (Si-PLS) coupled with Near-infrared (NIR) spectroscopy. These algorithms were independently applied to select informative spectral variables and improve the prediction of green tea components. Results showed that NIR combined with SA-PLS and Si-PLS had a strong correlation coefficient with the wet-chemical methods for predicting epigallocatechin gallate (Rp2 = 0.97); epigallocatechin (Rp2 = 0.97); epicatechin gallate (Rp2 = 0.96); epicatechin (Rp2 = 0.91); catechin (Rp2 = 0.98); caffeine (Rp2 = 0.96); theanine (Rp2 = 0.93); and antioxidant capability (Rp2 = 0.80) in green tea. Our results revealed the potential utilization of NIR spectroscopy coupled with SA-PLS and Si-PLS algorithms as an effective and robust technique to simultaneously predict active constituents and antioxidant capability of green tea.

Nondestructive determination of lignin content in Korla fragrant pear based on near-infrared spectroscopy

The lignin content is a key index affecting the quality of Korla fragrant pear. This study applied a method of near-infrared diffuse reflectance spectroscopy combined with chemometric method to establish a predictive model of lignin content. Among them, moving average, baseline, standard normal variate transformations and multiplicative scatter correction were used to preprocess the collected near-infrared spectra. then, the backward interval partial least-squares, synergy interval partial least squares and uninformative variable elimination algorithms were used to select characteristic wavelengths from the whole spectral range to establish a partial least-squares prediction model of lignin content. The partial least-squares model based on uninformative variable elimination selected characteristic wavelengths was simplified, and the prediction of lignin content in Korla pear was more accurate. Multiple determination coefficient values, standard prediction errors and residual prediction errors for prediction were 0.87, 1.36 and 2.03%, respectively. Compared with partial least squares based on the whole spectral range and partial least squares based on Uninformative variable elimination selection, the number of characteristic wavelength variables was reduced from 289 to 51, thereby improving the prediction accuracy of the model. The experimental results showed that the uninformative variable elimination by partial least-squares model was the best model. The combination of near-infrared spectroscopy and uninformative variable elimination optimization can effectively determine the lignin content in Korla fragrant pear. The near-infrared diffuse reflectance spectroscopy proved to be a good tool for nondestructive determination of lignin content in Korla fragrant pear. The selection of characteristic wavelengths and appropriate pretreatment methods can improve the accuracy of near-infrared spectroscopy in actual detection.

Rapid Determination of Total Polar Compounds in Frying Oil Using ATR‐FTIR Combined with Extended Partial Least Squares Regression

AbstractA method of rapidly determining the total polar compounds (TPCs) in frying oils using attenuated total reflectance‐Fourier transform infrared spectroscopy combined with partial least squares (PLS) regression is developed. Oils of various types and geographic origins are used to ensure that the proposed model is robust. The first derivative spectrum is selected as the spectral processing method. The interval PLS, forward interval PLS, and backward interval PLS algorithms are compared in terms of their performance. A correlation coefficient (R2) of 0.9942, a root mean square error of calibration (RMSEC) of 1.1, a root mean square error of prediction (RMSEP) of 2.30, a residual predictive deviation (RPD) of 4.1, and a limit of detection (LOD) of 1.65% are obtained by the fiPLS33 model with fewer latent variables and a lower spectral interval number. In addition, sub‐models using a single type of oil showed higher performance (R2 0.9957–0.9998, RMSEC 0.12–0.92, RMSEP 0.79–1.58, RPD 4.79–9.64, LOD 0.66–1.26%) than the general model. The TPC models developed are accurate, stable, and adaptable, and they can be used to analyze general frying oil samples quickly, regardless of the oil type, and to analyze samples of specific oil types accurately.Practical applications: The content of TPCs is an important indicator of whether the oil has been overused and whether it will be harmful during the frying process. However, traditional chemical methods are time‐consuming, and they have not been used to determine large‐sized samples. In addition, due to a lack of regional optimization, most studies on determining TPCs with FTIR give unsatisfactory model performance. A general TPC model that incorporates several oil types and regional optimization is expected to improve prediction performance. Therefore, the proposed method represents a rapid and accurate tool for measuring TPCs in edible fats and oils.

Rapid spectral analysis of agro-products using an optimal strategy: dynamic backward interval PLS-competitive adaptive reweighted sampling.

A novel strategy of variable selection approach named dynamic backward interval partial least squares-competitive adaptive reweighted sampling (DBiPLS-CARS) was proposed in this study. Near-infrared data sets of three different agro-products, namely corn, crop processing lamina, and plant leaf samples, were collected to investigate the performance of the proposed method. Weak relevant variables were first removed by DBiPLS and a refined selection of the remaining variables was then conducted by CARS. The Monte Carlo uninformative variable elimination (MCUVE) was used as a classical beforehand uninformative variable elimination method for comparison. Results showed that DBiPLS can select informative variables more continuously than MCUVE. Some synergistic variables which may be omitted by MCUVE can be retained by DBiPLS. By contrast, MCUVE can hardly avoid the disturbance of certain weak relevant variables as a result of its calculation based on the full spectrum regression. Therefore, DBiPLS exhibited the advantage of removing the weak relevant variables before CARS, and simultaneously improved the prediction performance of CARS.

Predicting macronutrients and energy content of snack products using FT-NIR analysis and chemometric techniques

Nowadays, the consumption of snack products is permanently increasing. Because of the growing trend of snack consumption, it is more and more difficult to guarantee the quality and safety of the products. Near infrared spectroscopy (NIRS) method, combined with chemometric techniques provide outstanding solutions, due to its rapidity and simple sample preparation. The objective of this study was to investigate the possibilities of using NIRS to predict fat, protein, carbohydrate, sugar and salt content of all in all 155 commercially available snack products from 25 countries. The prediction models were performed using partial least squares regression (PLSR) with different spectral pre-processing methods. Different pre-processing methods proved to be the best to predict the five macronutrients, however, the final models showed good accuracy |R2/Q2 > 0.94/0.82|. The energy content of the samples was calculated from the measured parameters and interval PLS regression was accomplished to improve prediction parameters. The methods developed are suitable for analyzing snacks made from single or mixed raw materials.

Non-Invasive Blood Glucose Estimation using Handheld Near Infra-Red Device

Non-invasive blood glucose measurement would ease everyday life of diabetic patients and may cut the cost involved in their treatments. This project aims at developing a non-invasive blood glucose measurement using NIR (near infrared) spectroscopic device. NIR spectra data and blood glucose levels were collected from 45 participants, resulting 90 samples (75 samples for calibration and 15 samples for testing) in this project. These samples were then used to develop a predictive model using Interval Partial Least Square (IPLS) regression method. The results obtained from this project indicate that the handheld micro NIR has potential use for rapid non-invasive blood glucose monitoring. The coefficient of determination (R 2 ) obtained for calibration/training and testing dataset are respectively 0.9 and 0.91.

Quantification of Jatropha methyl biodiesel in mixtures with diesel using mid-infrared spectrometry and interval variable selection methods

The use of mid-infrared spectrometry with horizontal attenuated total reflectance and Fourier transform, along with methods of variable selection by intervals, that is, interval partial least-squares (iPLS), backward interval partial least-squares (biPLS), and synergy interval partial least-squares (siPLS), were evaluated in order to quantify the Jatropha methyl biodiesel content in mixtures with diesel in the range from 0.25 to 30.00% (v/v). The spectral data were obtained in quintuplicate and corrected using the baseline technique. The constructed variable selection models were compared with the global partial least-squares (PLS) model by applying the F test to the root mean square error of prediction (RMSEP) value. The model with the best predictive capacity was iPLS24 (i.e., dividing the full spectrum into 24 equidistant intervals) constructed in the region from 650 to 750 cm−1 with 101 variables. The parameters obtained were the root mean square error of calibration (RMSEC) of 0.22%, the root mean square error of cross-validation (RMSECV) of 0.25%, RMSEP of 0.24%, limit of detection (LD) of 0.40% (v/v) and limit of quantification (LQ) of 0.13% (v/v). Therefore, the interval variable selection method was efficient in obtaining a spectral region that provided better predictive capacity than the global PLS model. In addition, with the selection of a smaller number of variables, it is possible to build portable equipment at a lower cost for use by fuel quality control agencies.

Fast determination of oxides content in cement raw meal using NIR spectroscopy combined with synergy interval partial least square and different preprocessing methods

Near infrared (NIR) diffuse reflectance spectroscopy, combined with chemometrics techniques was proposed as a fast and accurate method applied for determination of main components (CaO, SiO2, Al2O3, Fe2O3) in cement raw meal. The analyses were implemented by correlation of NIR spectra with reference method made by X-ray Fluorescence (XRF), using different regression methods such as partial least square (PLS), interval PLS (iPLS) and synergy interval PLS (siPLS) with multiple pretreatment methods. The optimal models obtained by using siPLS algorithm showed root-mean-square-error-of-prediction (RMSEP) from 0.0379 to 0.1715, the correlation coefficient (Rp) from 0.7294 to 0.9304 and average prediction error from 0.03% to 0.13%. The proposed method is cheaper and safer, and it takes much less time (3–5 min) than the XRF method (40–60 min) for each measurement. Our results demonstrate that NIR diffuse reflectance spectroscopy technology can be applied to quantitative determination of the four oxides in cement raw meal.

Fast determination of oxides content in cement raw meal using NIR-spectroscopy and backward interval PLS with genetic algorithm.

Determining oxides content in cement raw meal with near infrared (NIR) spectroscopy, associated with partial least square (PLS) regression, is fast and potential for cement industry to realize cement raw material proportioning control. However, it has hardly been studied. Backward interval PLS (biPLS) with genetic algorithm (GA-biPLS) were applied to select characteristic variables closely related to the concentration of oxide of interest to establish calibration model. The optimal GA-biPLS models showed that the determination coefficient (Rp2) and root mean square error of prediction (RMSEP) were 0.8857 and 0.0994 for CaO, 0.8718 and 0.1044 for SiO2, 0.7417 and 0.0693 for Al2O3, 0.5404 and 0.0387 for Fe2O3, correspondingly. These results indicate that GA-biPLS can select less variables with better prediction performance by comparison with PLS and biPLS, the NIR spectroscopy combined with GA-biPLS algorithm is a fast, accurate and reliable alternative method for determination of oxides content in cement raw meal.

Determination of API Gravity and Total and Basic Nitrogen Content by Mid- and Near-Infrared Spectroscopy in Crude Oil with Multivariate Regression and Variable Selection Tools

American Petroleum Institute (API) gravity is an important parameter in the crude oil industry and the nitrogen compounds are related to the toxic effects of the oil in refineries and the environment. In this paper, 194 crude oil samples with API gravities ranging from 11.4 to 57.5 were used for the purpose of estimating the physicochemical properties: API gravity, total nitrogen content (TNC) and basic nitrogen content (BNC). Initially, infrared spectra in the mid and near regions (MIR and NIR) were collected, then full-spectral partial least squares (PLS) and the orthogonal projections to latent structures (OPLS) chemometric models were developed and validated, as well as models using interval PLS (iPLS), synergy interval PLS (siPLS) and competitive adaptive reweighted sampling PLS (CARSPLS) as variable selection tools. For API gravity and TNC, the best calibration technique is the NIR CARSPLS with a root mean square error of prediction (RMSEP) values of 0.9 and 0.0275 wt%, respectively. For BNC, the best technique is MIR siPLS with a prediction error of 0.0134 wt%. The results were validated based on the evaluation of the figures of merit, a statistical evaluation of the accuracy, characterization of the systematic error and measurement for errors in the residues. The results were satisfactory considering the high variability of the data and the diversity of the samples, demonstrating suitable applicability for practical analysis.