GA-ACE: Alternating conditional expectations regression with selection of significant predictors by genetic algorithms

The non-linear regression technique known as alternating conditional expectations (ACE) method is only applicable when the number of objects available for calibration is considerably greater than the number of considered predictors. Alternating conditional expectations regression with selection of significant predictors by genetic algorithms (GA-ACE), the non-linear regression technique presented here, is based on the ACE algorithm but introducing several modifications to resolve the applicability limitations of the original ACE method, thus facilitating the practical implementation of a very interesting calibration tool. In order to overcome the lack of reliability displayed by the original ACE algorithm when working on data sets characterized by a too large number of variables and prior to the development of the non-linear regression model, GA-ACE applies genetic algorithms as a variable selection technique to select a reduced subset of significant predictors able to accurately model and predict a considered variable response. Furthermore, GA-ACE actually provides two alternative application approaches, since it allows either the performance of prior data compression computing a number of principal components to be subsequently subjected to GA-selection, or working directly on original variables.In this study, GA-ACE was applied to two real calibration problems, with a very low observation/variable ratio (NIR data), and the results were compared with those obtained by several linear regression techniques usually employed. When using the GA-ACE non-linear method, notably improved regression models were developed for the two response variables modeled, with root mean square errors of the residuals in external prediction (RMSEP) equal to 11.51 and 6.03% for moisture and lipid contents of roasted coffee samples, respectively. The improvement achieved by applying the new non-linear method introduced is even more remarkable taking into account the results obtained with the best performance linear method (IPW-PLS) applied to predict the studied responses (14.61 and 7.74% RMSEP, respectively).     

Multicomponent analysis of electrochemical signals in the wavelet domain

Successful applications of multivariate calibration in the field of electrochemistry have been recently reported, using various approaches such as multilinear regression (MLR), continuum regression, partial least squares regression (PLS) and artificial neural networks (ANN). Despite the good performance of these methods, it is nowadays accepted that they can benefit from data transformations aiming at removing baseline effects, reducing noise and compressing the data. In this context the wavelet transform seems a very promising tool. Here, we propose a methodology, based on the fast wavelet transform, for feature selection prior to calibration. As a benchmark, a data set consisting of lead and thallium mixtures measured by differential pulse anodic stripping voltammetry and giving seriously overlapped responses has been used. Three regression techniques are compared: MLR, PLS and ANN. Good predictive and effective models are obtained. Through inspection of the reconstructed signals, identification and interpretation of significant regions in the voltammograms are possible.     

Efficient use of pure component and interferent spectra in multivariate calibration

Partial Least Squares (PLS) is by far the most popular regression method for building multivariate calibration models for spectroscopic data. However, the success of the conventional PLS approach depends on the availability of a ‘representative data set’ as the model needs to be trained for all expected variation at the prediction stage. When the concentration of the known interferents and their correlation with the analyte of interest change in a fashion which is not covered in the calibration set, the predictive performance of inverse calibration approaches such as conventional PLS can deteriorate. This underscores the need for calibration methods that are capable of building multivariate calibration models which can be robustified against the unexpected variation in the concentrations and the correlations of the known interferents in the test set. Several methods incorporating ‘a priori’ information such as pure component spectra of the analyte of interest and/or the known interferents have been proposed to build more robust calibration models. In the present study, four such calibration techniques have been benchmarked on two data sets with respect to their predictive ability and robustness: Net Analyte Preprocessing (NAP), Improved Direct Calibration (IDC), Science Based Calibration (SBC) and Augmented Classical Least Squares (ACLS) Calibration. For both data sets, the alternative calibration techniques were found to give good prediction performance even when the interferent structure in the test set was different from the one in the calibration set. The best results were obtained by the ACLS model incorporating both the pure component spectra of the analyte of interest and the interferents, resulting in a reduction of the RMSEP by a factor 3 compared to conventional PLS for the situation when the test set had a different interferent structure than the one in the calibration set.     

Multivariate calibration leverages and spectral F‐ratios via the filter factor representation

Diagnostics are fundamental to multivariate calibration (MC). Two common diagnostics are leverages and spectral F‐ratios and these have been formulated for many MC methods such as partial least square (PLS), principal component regression (PCR) and classical least squares (CLS). While these are some of the most common methods of calibration in analytical chemistry, ridge regression is also common place and yet spectral F‐ratios have not been developed for it. Noting that ridge regression is a form of Tikhonov regularization (TR) and using the unifying filter factor representation for MC, this paper develops the filter factor form of leverages and spectral F‐ratios. The approach is applied to a spectral data set to demonstrate computational speed‐up advantages and ease of implementation for the filter factor representation. Copyright © 2010 John Wiley & Sons, Ltd.     

Practical guidelines for reporting results in single- and multi-component analytical calibration: A tutorial

Practical guidelines for reporting analytical calibration results are provided. General topics, such as the number of reported significant figures and the optimization of analytical procedures, affect all calibration scenarios. In the specific case of single-component or univariate calibration, relevant issues discussed in the present Tutorial include: (1) how linearity can be assessed, (2) how to correctly estimate the limits of detection and quantitation, (2) when and how standard addition should be employed, (3) how to apply recovery studies for evaluating accuracy and precision, and (4) how average prediction errors can be compared for different analytical methodologies. For multi-component calibration procedures based on multivariate data, pertinent subjects here included are the choice of algorithms, the estimation of analytical figures of merit (detection capabilities, sensitivity, selectivity), the use of non-linear models, the consideration of the model regression coefficients for variable selection, and the application of certain mathematical pre-processing procedures such as smoothing.     

Using the L1 norm to select basis set vectors for multivariate calibration and calibration updating

With projection based calibration approaches, such as partial least squares (PLS) and principal component regression (PCR), the calibration space is spanned by respective basis vectors (latent vectors). Up to rank k basis vectors are formed where k ≤ min(m,n) with m and n denoting the number of calibration samples and measured variables. The user needs to decide how many and which respective basis vectors (tuning parameters). To avoid the second issue, basis vectors are selected top‐down starting with the first and sequentially adding until model criteria are satisfied. Ridge regression (RR) avoids the issues by using the full set of basis vectors. Another approach is to select a subset from the total available. The presented work develops a process based on the L₁ vector norm to select basis vectors. Specifically, the L₁ norm is used to select singular value decomposition (SVD) basis set vectors for PCR (LPCR). Because PCR, PLS, RR, and others can be expressed as linear combination of the SVD basis vectors, the focus is on selection and comparison using the SVD basis set. Results based on respective tuning parameter selections and weights applied to the SVD basis vectors for LPCR, top‐down PCR, correlation PCR (CPCR), PLS, and RR are compared for calibration and calibration updating using spectroscopic data sets. The methods are found to predict equivalently. In particular, the L₁ norm produces similar results to those obtained by the well‐studied CPCR process. Thus, the new method provides a different theoretical framework than CPCR for selecting basis vectors. Copyright © 2016 John Wiley & Sons, Ltd.     

Different approaches to multivariate calibration of nonlinear sensor data

In this study, different approaches to the multivariate calibration of the vapors of two refrigerants are reported. As the relationships between the time-resolved sensor signals and the concentrations of the analytes are nonlinear, the widely used partial least-squares regression (PLS) fails. Therefore, different methods are used, which are known to be able to deal with nonlinearities present in data. First, the Box–Cox transformation, which transforms the dependent variables nonlinearly, was applied. The second approach, the implicit nonlinear PLS regression, tries to account for nonlinearities by introducing squared terms of the independent variables to the original independent variables. The third approach, quadratic PLS (QPLS), uses a nonlinear quadratic inner relationship for the model instead of a linear relationship such as PLS. Tree algorithms are also used, which split a nonlinear problem into smaller subproblems, which are modeled using linear methods or discrete values. Finally, neural networks are applied, which are able to model any relationship. Different special implementations, like genetic algorithms with neural networks and growing neural networks, are also used to prevent an overfitting. Among the fast and simpler algorithms, QPLS shows good results. Different implementations of neural networks show excellent results. Among the different implementations, the most sophisticated and computing-intensive algorithms (growing neural networks) show the best results. Thus, the optimal method for the data set presented is a compromise between quality of calibration and complexity of the algorithm.Electronic Supplementary Material Supplementary material is available for this article at     

Non-linear calibration models for near infrared spectroscopy

Different calibration techniques are available for spectroscopic applications that show nonlinear behavior. This comprehensive comparative study presents a comparison of different nonlinear calibration techniques: kernel PLS (KPLS), support vector machines (SVM), least-squares SVM (LS-SVM), relevance vector machines (RVM), Gaussian process regression (GPR), artificial neural network (ANN), and Bayesian ANN (BANN). In this comparison, partial least squares (PLS) regression is used as a linear benchmark, while the relationship of the methods is considered in terms of traditional calibration by ridge regression (RR). The performance of the different methods is demonstrated by their practical applications using three real-life near infrared (NIR) data sets. Different aspects of the various approaches including computational time, model interpretability, potential over-fitting using the non-linear models on linear problems, robustness to small or medium sample sets, and robustness to pre-processing, are discussed. The results suggest that GPR and BANN are powerful and promising methods for handling linear as well as nonlinear systems, even when the data sets are moderately small. The LS-SVM is also attractive due to its good predictive performance for both linear and nonlinear calibrations.     

An alternative quadrilinear decomposition algorithm for four-way calibration with application to analysis of four-way fluorescence excitation–emission–pH data array

A novel quadrilinear decomposition algorithm for four-way calibration (third-order tensor calibration), which was called as regularized self-weighted alternating quadrilinear decomposition (RSWAQLD), has been developed in this work. It originates from the alternating trilinear decomposition (ATLD) algorithm, inherits the philosophy behind self-weighting operation from the self-weighted alternating trilinear decomposition (SWATLD) algorithm. The RSWAQLD algorithm is based on a nearby least-squares scheme, in which two extra terms are added to each loss function, making it more stable and flexible. Experiment shows that RSWAQLD has the features of fast convergence and being insensitive to the excess estimated factors in the model. Owing to its unique optimizing approach, RSWAQLD is much more efficient than four-way PARAFAC. Moreover, the performance of RSWAQLD is quit stable as the number of factors used in calculation varies (as long as it is no less than the true number of factors). Such a feature will simplify the analysis of four-way data arrays, since it is unnecessary to spend a lot of time and effort on accurately determining the appropriate number of factors in the matrix. In addition, the result of four-way fluorescence excitation–emission–pH data, as well as that of simulated data, illustrated that RSWAQLD can not only remain the “higher-order advantage” but also provide a satisfying result even in high collinear systems.     

Influence of data pre-processing on the quantitative determination of the ash content and lipids in roasted coffee by near infrared spectroscopy

In near-infrared (NIR) measurements, some physical features of the sample can be responsible for effects like light scattering, which lead to systematic variations unrelated to the studied responses. These errors can disturb the robustness and reliability of multivariate calibration models. Several mathematical treatments are usually applied to remove systematic noise in data, being the most common derivation, standard normal variate (SNV) and multiplicative scatter correction (MSC). New mathematical treatments, such as orthogonal signal correction (OSC) and direct orthogonal signal correction (DOSC), have been developed to minimize the variability unrelated to the response in spectral data. In this work, these two new pre-processing methods were applied to a set of roasted coffee NIR spectra. A separate calibration model was developed to quantify the ash content and lipids in roasted coffee samples by PLS regression. The results provided by these correction methods were compared to those obtained with the original data and the data corrected by derivation, SNV and MSC. For both responses, OSC and DOSC treatments gave PLS calibration models with improved prediction abilities (4.9 and 3.3% RMSEP with corrected data versus 7.1 and 8.3% RMSEP with original data, respectively).     

Determination of carbamazepine in serum and pharmaceutical preparations using immobilization on a nylon support and fluorescence detection

A novel approach is presented for the spectrofluorimetric determination of the powerful anticonvulsant carbamazepine and its main metabolite in human serum. The strategy consists in the support of both compounds on a nylon membrane, and their subsequent determination through a solid-surface fluorescence methodology combined with a suitable chemometric analysis. The novelty of the present method lies in the fact that while carbamazepine does not fluoresce neither in solution nor supported on a variety of surfaces, significant emission signals are observed when it is supported on the nylon matrix, a property which has not been previously exploited by analysts. Multivariate calibration analysis was performed on three-way excitation-emission matrix data. The algorithms applied were: parallel factor analysis (PARAFAC), self-weighted alternating trilinear decomposition (SWATLD) and N-way partial least-squares regression (N-PLS). The results were compared with two-way calibration data analysed with partial least-squares regression (PLS-1). The methodology is highly specific, and it appears to be suitable for the routine monitoring of serum concentrations in patients receiving chronic therapy. In addition, the technique was satisfactorily applied to the determination of carbamazepine in pharmaceutical formulations.     

Qualitative and quantitative analysis of oxytetracycline by near-infrared spectroscopy

Near-infrared (NIR) spectroscopy, in combination with chemometrics, enable the analysis of raw materials without time-consuming sample preparation methods. The aim of our work was to estimate critical parameters in the analytical specification of oxytetracycline, and consequently the development of a method for quantification and qualification of these parameters by NIR spectroscopy. A Karl Fischer (K.F.) titration to determine the water content, a colorimetric assay method, and Fourier transform-infrared (FT-IR) spectroscopy to identify the oxytetracycline base, were used as reference methods, respectively. Multivariate calibration was performed on NIR spectral data using principal component analysis (PCA), partial least-squares (PLS 1) and principal component regression (PCR) chemometric methods. Multivariate calibration models for NIR spectroscopy have been developed. Using PCA and the Soft Independent Modelling of Class Analogy (SIMCA) approach, we established the cluster model for the determination of sample identity. PLS 1 and PCR regression methods were applied to develop the calibration models for the determination of water content and the assay of the oxytetracycline base. Comparing the PLS and PCR regression methods we found out that the PLS is better established by NIR, especially as the spectroscopic data (NIR spectra) are highly collinear and there are many wavelengths due to non-selective wavelengths. The calibration models for NIR spectroscopy are convenient alternatives to the colorimetric method and to the K.F. method, as well as to FT-IR spectroscopy, in the routine control of incoming material.     

Discrimination of Beers by Cyclic Voltammetry Using a Single Carbon Screen-printed Electrode

A fast, simple and costless methodology without sample pre-treatment is proposed for the discrimination of beers. It is based on cyclic voltammetry (CV) using commercial carbon screen-printed electrodes (SPCE) and includes a correction of the signals measured with different SPCE units. Data are submitted to partial least squares discriminant analysis (PLS−DA) and support vector machine discriminant analysis (SVM−DA), which allow a reasonable classification of the beers. Also, CV data from beers can be used to predict their alcoholic degree by partial least squares (PLS) and artificial neural networks (ANN). In general, non-linear methods provide better results than linear ones.     

Least-squares support vector machines and near infrared spectroscopy for quantification of common adulterants in powdered milk

This paper proposes the use of the least-squares support vector machine (LS-SVM) as an alternative multivariate calibration method for the simultaneous quantification of some common adulterants (starch, whey or sucrose) found in powdered milk samples, using near-infrared spectroscopy with direct measurements by diffuse reflectance. Due to the spectral differences of the three adulterants a nonlinear behavior is present when all groups of adulterants are in the same data set, making the use of linear methods such as partial least squares regression (PLSR) difficult. Excellent models were built using LS-SVM, with low prediction errors and superior performance in relation to PLSR. These results show it possible to built robust models to quantify some common adulterants in powdered milk using near-infrared spectroscopy and LS-SVM as a nonlinear multivariate calibration procedure.     

Simultaneous determination of three fluoroquinolones by linear sweep stripping voltammetry with the aid of chemometrics

A linear sweep stripping voltammetric (LSSV) method has been researched and developed for simultaneous quantitative determination of mixtures of three antibiotic drugs, ofloxacin, norfloxacin and ciprofloxacin. It relies on reductive reaction of the antibiotics at a mercury electrode in a Britton-Robinson buffer (pH 3.78). The voltammograms of these three compounds overlap strongly, and show non-linear character. Thus, it is difficult to analyse the compounds individually in their mixtures. In this work, chemometrics methods such as classical least squares (CLS), principal component regression (PCR), partial least squares (PLS) and radial basis function-artificial neural networks (RBF-ANN) were applied for the simultaneous determination of these compounds. The prediction performance of the calibration models constructed on the basis of these methods was compared. It was shown that satisfactory quantitative results were obtained with the use of the RBF-ANN calibration model relative prediction error (RPE_T) of 8.1% and an average recovery of 101%. This method is able to accommodate non-linear data quite well. The proposed analytical method based on LSSV was applied for the analysis of ofloxacin, norfloxacin and ciprofloxacin antibiotics in bird feedstuffs and their spiked samples, as well as in eye drops with satisfactory results.     

叠加多元校正分析

基于多模型(模型融合)建模的思想,开发了两种新的叠加多元校正分析算法:叠加PCR(PLS)多元校正分析和叠加移动窗口PCR(PLS)多元校正分析。与一般的多模型建模方法不同的是其通过赋予光谱数据中的不同部分不同权重叠加子多元校正模型。因此,其可以通过权重调节或选择变量。在消除光谱数据中常见的冗余信息的同时,避免信息遗漏的缺点,并最终提高模型的稳健性,简化了模型。对于这两个新的算法,尽管其具体步骤不同,但仍取得了相似的预测结果。本文通过两套近红外光谱文献数据计算验证了这两个新方法的优越性。     

三维荧光二阶校正法用于未知混合物中苯酚含量的测定

将三维荧光光谱技术与秩消失因子分析、广义秩消失因子分析和交替三线性分解3种二阶校正方法相结合,建立了测定未知混合物中苯酚含量的三维荧光二阶校正新方法。设定在激发波长240~280 nm和发射波长280~360 nm范围内测定未知混合物中苯酚的三维荧光光谱,构建三维响应数据阵,运用基于三线性分解的二阶校正算法进行解析。结果表明,当模拟样品的组分数为2时,秩消失因子分析、广义秩消失因子分析和交替三线性分解3种方法测定苯酚的预测均方根误差分别为0.33,1.18和0.15,平均回收率分别为101.6%,115.6%和101.9%;当组分数为3时,3种方法的预测均方根误差则分别为1.61,1.80和0.51,平均回收率分别为134.2%,133.9%和107.1%;将其分别应用于实际样品中苯酚的测定,结果满意,且交替三线性分解法的测定结果优于秩消失因子分析法和广义秩消失因子分析法。     

Boosted regression trees, multivariate adaptive regression splines and their two-step combinations with multiple linear regression or partial least squares to predict blood-brain barrier passage: a case study

The use of some unconventional non-linear modeling techniques, i.e. classification and regression trees and multivariate adaptive regression splines-based methods, was explored to model the blood-brain barrier (BBB) passage of drugs and drug-like molecules. The data set contains BBB passage values for 299 structural and pharmacological diverse drugs, originating from a structured knowledge-based database. Models were built using boosted regression trees (BRT) and multivariate adaptive regression splines (MARS), as well as their respective combinations with stepwise multiple linear regression (MLR) and partial least squares (PLS) regression in two-step approaches. The best models were obtained using combinations of MARS with either stepwise MLR or PLS. It could be concluded that the use of combinations of a linear with a non-linear modeling technique results in some improved properties compared to the individual linear and non-linear models and that, when the use of such a combination is appropriate, combinations using MARS as non-linear technique should be preferred over those with BRT, due to some serious drawbacks of the BRT approaches.     

人工神经网络用于近红外光谱测定柴油闪点

采用主成分－人工神经网络对不同留程柴油的近红外光谱进行校正,预测其闪点。采用监控集控制网络训练过程,以避免过训练。探讨了人工神经网络（ＡＮＮ）、直接线性连接人工神经网络（ＬＡＮＮ）的校正效果,并与局部权重回归（ＬＷＲ）、主成分回归（ＰＣＲ）及偏最小二乘（ＰＬＳ）等校正方法进行了比较,认为人工神经及直接线性关联的较好手段。