PLS Method Research Articles

QSPR prediction of thermal decomposition property of non-vinyl polymers having α-amino acids moieties

In the present work, a quantitative structure–property relationship (QSPR) treatment of temperature of five percent of decomposition (T 5) of a number of totally 30 optically active polymers was performed by means of a genetic algorithm-based partial least squares (GA–PLS) and artificial neural network (ANN). Suitable set of molecular descriptors were calculated by dragon package and the important descriptors were selected by GA–PLS methods. These descriptors were served as inputs to generate ANN. After optimization and training of the networks, they were used for the calculation of T 5 for the validation set. By comparing of the results obtained from PLS and ANN models, it can be seen that statistical parameters (Fisher ratio, correlation coefficient, and standard error) of the ANN model are better than PLS one, which indicates that nonlinear model can simulate the relationship between the structural descriptors and T 5 of the investigated macromolecules more accurately.

A Nonlinear System Identification Method Based on Local Linear PLS Method

A Nonlinear System Identification Method Based on Local Linear PLS Method

An Empirical Study on the Determinants of Foreign Subsidiaries' Marketing Knowledge Sharing Activity and its Effect on Marketing Knowledge Sharing Satisfaction

This paper attempts to examine how the relationship factors of peer foreign subsidiaries in MNEs affect marketing knowledge sharing activity between them and whether the activity guarantees the quality of the shared knowledge and leads to marketing knowledge sharing satisfaction. The research model were tested with the data collected from 113 foreign subsidiaries of Korean MNEs publicly listed in the Korean Stock Exchange, using PLS method. The empirical results show that mutual dependence and trust between them were found to positively influence the extent to which they are actively involved in marketing knowledge sharing and more satisfaction is expected in foreign subsidiaries which share marketing knowledge actively. However, active sharing of marketing knowledge does not always satisfy the parties. They are satisfied only when the shared knowledge is of good quality. The result shows more active sharing of marketing knowledge improves the quality of the shared knowledge owing to a kind of experience and a learning effect and consequently, the subsidiaries are more satisfied.

Business Intelligence Practices for Exporting SMEs

The aim of this paper is to study business intelligence practices of exporting SME and their relationship to the level of international commitment. The results show that managers’ sensitivity, influenced by the level of international commitment, plays a central role in the organization of their intelligence practices. The research focused on Quebec SME exporters and used the PLS method to verify research assumptions.

QSPR STUDY OF THE SOLUTE POLARITY PARAMETER IN REVERSED-PHASE LIQUID CHROMATOGRAPHY USING PARTIAL LEAST SQUARES AND ARTIFICIAL NEURAL NETWORK

A quantitative structure-property relationship (QSPR) study based on artificial neural network (ANN) was carried out for the prediction of the solute polarity parameters of a set of 146 compounds of a very different chemical nature in reversed-phase liquid chromatography (RPLC). The genetic algorithm-partial least squares (GA-PLS) method was applied as a variable selection tool. A PLS method was used to select the best descriptors and the selected descriptors were used as input neurons in neural network model. These descriptors are: topological electronic index (), total hybridization components of the molecular dipole (μh), total dipole moment of the molecule (μ), PPSA-3 atomic charge weighted PPSA (PPSA-3), Kier & Hall index, order 0 (0χ), and Molecular volume (MV). The results obtained showed the ability of developed artificial neural network to predict the solute polarity parameter of various compounds. Moreover, results reveal the superiority of the ANN over the PLS model.

A multivariate analysis of protein microarrays for signature selection profiles

A multivariate analysis of protein microarrays for signature selection profiles

Fourier transform near-infrared spectroscopy (FT-NIRS) application to estimate Brazilian soybean [Glycine max (L.) Merril] composition

This study examined the ability of near-infrared reflectance spectroscopy method (FT-NIRS) and multivariate calibration to estimate the concentration of moisture, protein, lipid, ash and carbohydrate of Brazilian soybeans. The spectra obtained in the range of 4000 to 10,000cm−1 were preprocessed by several combinations of mathematical treatments: MSC (multiplicative scatter correction), SNV (standard normal variate) or first and second derivative and all data were mean centered before the calibration, for which was used the PLS method (partial least squares). The best calibration models found in this study were the ones used to determine protein and moisture contents (R2=0.81, RMSEP=1.61% and R2=0.80, RMSEC=1.55%, respectively). However, the technique shows high predictability for all parameters, including lipids, ashes and carbohydrates, with RMSECV of 0.40 to 2.30% and RMSEP of 0.38 to 3.71%. This result shows the viability of using NIR in controlling the quality parameters of soybeans.

Strategies for multivariate modeling of moisture content in freeze-dried mannitol-containing products by near-infrared spectroscopy

Accurate determination of residual moisture content of a freeze-dried (FD) pharmaceutical product is critical for prediction of its quality. Near-infrared (NIR) spectroscopy is a fast and non-invasive method routinely used for quantification of moisture. However, several physicochemical properties of the FD product may interfere with absorption bands related to the water content. A commonly used stabilizer and bulking agent in FD known for variation in physicochemical properties, is mannitol. To minimize this physicochemical interference, different approaches for multivariate correlation between NIR spectra of a FD product containing mannitol and the corresponding moisture content measured by Karl Fischer (KF) titration have been investigated. A novel method, MIPCR (Main and Interactions of Individual Principal Components Regression), was found to have significantly increased predictive ability of moisture content compared to a traditional PLS approach. The philosophy behind the MIPCR is that the interference from a variety of particle and morphology attributes has interactive effects on the water related absorption bands. The transformation of original wavelength variables to orthogonal scores gives a new set of variables (scores) without covariance structure, and the possibility of inclusion of interaction terms in the further modeling. The residual moisture content of the FD product investigated is in the range from 0.7% to 2.6%. The mean errors of cross validated prediction of models developed in the investigated NIR regions were reduced from a range of 24.1-27.6% for traditional PLS method to 15.7-20.5% for the MIPCR method. Improved model quality by application of MIPCR, without the need for inclusion of a large number of calibration samples, might increase the use of NIR in early phase product development, where availability of calibration samples is often limited.

A QSRR Modeling of Hazardous Psychoactive Designer Drugs Using GA-PlS and L-M ANN

The hazardous psychoactive designer drugs are compounds in which part of the molecular structure of a stimulant or narcotic has been modified. A quantitative structure-retention relationship (QSRR) study based on a Levenberg-Marquardt artificial neural network (L-M ANN) was carried out for the prediction of the capacity factor (k′) of hazardous psychoactive designer drugs that contain Tryptamine, Phenylethylamine and Piperazine. The genetic algorithm-partial least squares (GA-PLS) method was used as a variable selection tool. A PLS method was used to select the best descriptors and the selected descriptors were used as input neurons in neural network model. For choosing the best predictive model from among comparable models, square correlation coefficient (R2) for the whole set is suggested to be a good criterion. Finally, to improve the results, structure-retention relationships were followed by nonlinear approach using artificial neural networks and consequently better results were obtained. Also this demonstrates the advantages of L-M ANN. This is the first research on the QSRR of the designer drugs using the GA-PLS and L-M ANN.

Detection of Drug Active Ingredients by Chemometric Processing of Solid-State NMR Spectrometry Data—The Case of Acetaminophen

This paper presents a preliminary study in building discriminant models from solid-state NMR spectrometry data to detect the presence of acetaminophen in over-the-counter pharmaceutical formulations. The dataset, containing 11 spectra of pure substances and 21 spectra of various formulations, was processed by partial least squares discriminant analysis (PLS-DA). The model found coped with the discrimination, and its quality parameters were acceptable. It was found that standard normal variate preprocessing had almost no influence on unsupervised investigation of the dataset. The influence of variable selection with the uninformative variable elimination by PLS method was studied, reducing the dataset from 7601 variables to around 300 informative variables, but not improving the model performance. The results showed the possibility to construct well-working PLS-DA models from such small datasets without a full experimental design.

Hyperspectral retrieval of phycocyanin in potable water sources using genetic algorithm–partial least squares (GA–PLS) modeling

Eagle Creek, Morse and Geist reservoirs, drinking water supply sources for the Indianapolis, Indiana, USA metropolitan region, are experiencing nuisance cyanobacterial blooms. Hyperspectral remote sensing has been proven to be an effective tool for phycocyanin (C-PC) concentration retrieval, a proxy pigment unique to cyanobacteria in freshwater ecosystems. An adaptive model based on genetic algorithm and partial least squares (GA–PLS), together with three-band algorithm (TBA) and other band ratio algorithms were applied to hyperspectral data acquired from in situ (ASD spectrometer) and airborne (AISA sensor) platforms. The results indicated that GA–PLS achieved high correlation between measured and estimated C-PC for GR (RMSE=16.3μg/L, RMSE%=18.2; range (R): 2.6–185.1μg/L), MR (RMSE=8.7μg/L, RMSE%=15.6; R: 3.3–371.0μg/L) and ECR (RMSE=19.3μg/L, RMSE%=26.4; R: 0.7–245.0μg/L) for the in situ datasets. TBA also performed well compared to other band ratio algorithms due to its optimal band tuning process and the reduction of backscattering effects through the third band. GA–PLS (GR: RMSE=24.1μg/L, RMSE%=25.2, R: 25.2–185.1μg/L; MR: RMSE=15.7μg/L, RMSE%=37.4, R: 2.0–135.1μg/L) and TBA (GR: RMSE=28.3μg/L, RMSE%=30.1; MR: RMSE=17.7μg/L, RMSE%=41.9) methods results in somewhat lower accuracy using AISA imagery data, which is likely due to atmospheric correction or radiometric resolution. GA–PLS (TBA) obtained an RMSE of 24.82μg/L (35.8μg/L), and RMSE% of 31.24 (43.5) between measured and estimated C-PC for aggregated datasets. C-PC maps were generated through GA–PLS using AISA imagery data. The C-PC concentration had an average value of 67.31±44.23μg/L in MR with a large range of concentration, while the GR had a higher average value 103.17±33.45μg/L.

Generalization of Quantification for PLS Correlation

This study proposes a quantification algorithm for a PLS method with several sets of variables. We called the quantification method for PLS with more than 2 sets of data a generalization. The basis of the quantification for PLS method is singular value decomposition. To derive the form of singular value decomposition in the data with more than 2 sets more easily, we used the constraint, a t a + b t b + c t c = 3 not a t a = 1, b t b = 1, and c t c = 1, for instance, in the case of 3 data sets. However, to prove that there is no difference, we showed it by the use of 2 data sets case because it is very complicate to prove with 3 data sets. The keys of the study are how to form the singular value decomposition and how to get the coordinates for the plots of variables and observations.

Partial least squares and random sample consensus in outlier detection

A novel outlier detection method in partial least squares based on random sample consensus is proposed. The proposed algorithm repeatedly generates partial least squares solutions estimated from random samples and then tests each solution for the support from the complete dataset for consistency. A comparative study of the proposed method and leave-one-out cross validation in outlier detection on simulated data and near-infrared data of pharmaceutical tablets is presented. In addition, a comparison between the proposed method and PLS, RSIMPLS, PRM is provided. The obtained results demonstrate that the proposed method is highly efficient.

Estimate Soil Organic Matter Content in the Heihe River Basin Using Hyperspectral Method

.Soil properties are very absent and sorely needed for many models. Conventional measurements are expensive and many samples are required. Hence it is difficult to quantitatively analysis them at a large-scale region. Hyperspectral methods are now used for the rapid characterization of a wide range of soil. In this paper, it was used to estimate organic matter content in soil of Heihe River basin. A total of 84 samples were collected and their reflected spectral were measured with ASD spectrometers. Then the organic matter were predicted using four forms of spectra and PLS method. The results show:1) the hyperspectral remote sensing can be used to predict soil organic matter content, and the precision can meet the requirement; 2) the best region of spectrum is 400~800nm in R, 1/R and Log (1/R) formation, and the FDR and FD(Log (1/R)) centered in 800nm based on the correlation coefficient and Variable Importance in Projection.

Estimate Soil Organic Matter Content in the Heihe River Basin Using Hyperspectral Method

.Soil properties are very absent and sorely needed for many models. Conventional measurements are expensive and many samples are required. Hence it is difficult to quantitatively analysis them at a large-scale region. Hyperspectral methods are now used for the rapid characterization of a wide range of soil. In this paper, it was used to estimate organic matter content in soil of Heihe River basin. A total of 84 samples were collected and their reflected spectral were measured with ASD spectrometers. Then the organic matter were predicted using four forms of spectra and PLS method. The results show:1) the hyperspectral remote sensing can be used to predict soil organic matter content, and the precision can meet the requirement; 2) the best region of spectrum is 400~800nm in R, 1/R and Log (1/R) formation, and the FDR and FD(Log (1/R)) centered in 800nm based on the correlation coefficient and Variable Importance in Projection.

The Role of the Convenience and the Subjective Norm in the Intention of Purchase Across Internet (B2C): an application in the hospitality industry

With the aim of studying those factors affecting the acceptance of a new distribution channel by the final consumer, we have analyzed the importance of acquiring by the convenience, subjective norm, usefulness and perceived ease of use on intention to buy on-line by the end user. The study of these variables was carried out specifically in the context of the Technology Acceptance Model and the Theory of Planned Behavior, whose theoretical basis is the Theory of Reasoned Action by causal model. For this we have conducted an on-line survey in Spain to a panel of Internet users during the months of December 2007 to January 2008. The final sample size was 190 Internet users. To test the causal model proposed in the research we used the PLS method, especially suitable for working with formative scales, as well as investigations in the field of TAM. The findings have revealed how convenience is the most important influence on the utility of using the Internet as a shopping channel, and one of the most important factors that, together with subjective norms, determines the purchase intentions through Internet. These findings are consistent with those found in previous researches in the relevant literature and had shown how convenience was the main motivating factor in online shopping. From these results it has been possible to propose several recommendations to improve online business management of enterprises in the sector. Key words: Convenience. Subjective norm. Online shopping. TAM. Hotel booking.

Hierarchical Bayes Modeling of the Customer Satisfaction Index

Customer Satisfaction Index has been developed in many countries including North America, Europe and Asia last decades, which are based on Americal Customer Satisfaction Index (ACSI) by the University of Michigan, where the latent factor “Customer Satisfaction” related to the customer loyalty is estimated by a covariance structural model with six factors generated from 17 question items and PLS method. They apply the identical structural model to all companies in order to measure the national and industrial indexes that are used to compare the services in different companies as well as industries. In this paper, by using the assumption that the same model must be applied to every company, we link the path coefficients of each company as the hierarchical regression model to estimate the structure for customer satisfaction across companies to show that, representing “communality” inside industry and “heterogeneity” outside industry, the hierarchical Bayes modeling produces more stable significant path coefficients. Moreover, our approach has the additional advantages. (i) The volume of information (number of survey data) can be augmented, (ii) The index can be constructed without additional surveys for new company (forecasting) and not-surveyed company (missing observations), (iii) When aggregating individual index of each company up to the industrial index and national index, the communality assumption could increase the stability of the macro index. [Service Science, ISSN 2164-3962 (print), ISSN 2164-3970 (online), was published by Services Science Global (SSG) from 2009 to 2011 as issues under ISBN 978-1-4276-2090-3.]

Fourier Transform Near Infrared Spectroscopy Analysis of Power Plant Coal Quality

Studied the correlation of near-infrared spectra data and six coal indices, found ash and calorific value has low correlations with spectra data; then use dynamic principal components PLS method predict the coal index; this method could predict Volatile and Hydrogen content better, however, the ability to predict Carbon content and Nitrogen content is lower. It is found that using reflection spectroscopy analyzes the coal need a strong energy source, because the color of coal is deep and reflection is very weak, this leads to noisy signals. Only by improving the energy source could solve the problem of poor spectra data fundamentally; spectral data mining cannot fundamentally improve the quality of data. The current near-infrared reflection spectroscopy common platform such as BRUKER is not suitable for coal analysis, analysis of coal to be better need to develop special near-infrared measuring instruments.

A non-linearized PLS model based on multivariate dominant factor for laser-induced breakdown spectroscopy measurements

A multivariate dominant factor based non-linearized PLS model is proposed. The intensities of different lines were taken to construct a multivariate dominant factor model, which describes the dominant concentration information of the measured species. In constructing such a multivariate model, non-linear transformation of multi characteristic line intensities according to the physical mechanisms of laser induced plasma spectrum were made, combined with linear-correlation-based PLS method, to model the nonlinear self-absorption and inter-element interference effects. This enables the linear PLS method to describe the non-linear relationship more accurately and provides the statistics-based PLS method with physical backgrounds. Moreover, a secondary PLS is applied utilizing the whole spectra information to further correct the model results. Experiments were conducted using standard brass samples. Taylor expansion was applied to make the nonlinear transformation describe the self-absorption effect of Cu. Then, line intensities of another two elements, Pb and Zn, were taken into account for inter-element interference. The proposed method shows a significant improvement when compared with conventional PLS model. Results also show that, even compared with the already-improved baseline dominant-factor-based PLS model, the present PLS model based on the multivariate dominant factor yields the same calibration quality (R2 = 0.999) while decreasing the RMSEP from 2.33% to 1.97%. The overall RMSE was also improved to 1.05% from 1.27%.

Dispersive Raman spectroscopy for the invitroidentification and quantification of injected vancomycin intra-vitreous

This work presents a study which aimed to quantify the intravitreally-injected vancomycin antibiotic using dispersive Raman spectroscopy and Partial Least Squares multivariate calibration. Eyes of New Zealand rabbits were injected with 5.0 mg vancomycin at different time intervals (8, 24 and 48 h groups). Distilled water was injected into the eyes of rabbits in the control group. The vitreous was then removed from each eyeball, placed into a quartz cuvette and submitted for dispersive Raman spectroscopy. The Raman equipment used a laser (80 mW, 830 nm), an imaging spectrograph and a CCD camera with exposure time of 50 s. A calibration curve was developed using the PLS method with vancomycin diluted in water in the range of 1.0–8.0 mg/ml. The spectrum of vancomycin was present in bands at the positions of C=C, amide I, CH3, amide III and vibrations of the carbon chain (C–C, aromatic ring breathing and C–O–C). The spectra of the injected vitreous bodies showed weak bands which were correlated with the pure drug. The PLS calibration model had a standard error of prediction of 0.8 mg/ml using one latent variable. The PLS showed changes in the predicted concentrations of the injected vitreous bodies depending on the time of injection, with an amount near zero for the water-injected group and ranging between 2 and 3 mg/ml for the drug-injected groups. Raman spectroscopy could be employed in the quantitative evaluation of intravitreally-injected vancomycin in future invivoapplications.