Variable selection in multivariate calibration based on clustering of variable concept

Machine learning methods in near infrared spectroscopy for predicting sensory traits in sweetpotatoes

It has been established that near infrared (NIR) spectroscopy has the potential of estimating sensory traits given the direct spectral responses that these properties have in the NIR region. In sweetpotato, sensory and texture traits are key for improving acceptability of the crop for food security and nutrition. Studies have statistically modelled the levels of NIR spectroscopy sensory characteristics using partial least squares (PLS) regression methods. To improve prediction accuracy, there are many advanced techniques, which could enhance modelling of fresh (wet and un-processed) samples or nonlinear dependence relationships. Performance of different quantitative prediction models for sensory traits developed using different machine learning methods were compared. Overall, results show that linear methods; linear support vector machine (L-SVM), principal component regression (PCR) and PLS exhibited higher mean R2 values than other statistical methods. For all the 27 sensory traits, calibration models using L-SVM and PCR has slightly higher overall R2 (x¯ = 0.33) compared to PLS (x¯ = 0.32) and radial-based SVM (NL-SVM; x¯= 0.30). The levels of orange color intensity were the best predicted by all the calibration models (R2 = 0.87 - 0.89). The elastic net linear regression (ENR) and tree-based methods; extreme gradient boost (XGBoost) and random forest (RF) performed worse than would be expected but could possibly be improved with increased sample size. Lower average R2 values were observed for calibration models of ENR (x¯ = 0.26), XGBoost (x¯ = 0.26) and RF (x¯ = 0.22). The overall RMSE in calibration models was lower in PCR models (X = 0.82) compared to L-SVM (x¯ = 0.86) and PLS (x¯ = 0.90). ENR, XGBoost and RF also had higher RMSE (x¯ = 0.90 - 0.92). Effective wavelengths selection using the interval partial least-squares regression (iPLS), improved the performance of the models but did not perform as good as the PLS. SNV pre-treatment was useful in improving model performance.

Machine Learning-Enabled NIR Spectroscopy. Part 2: Workflow for Selecting a Subset of Samples from Publicly Accessible Data

An increasingly large dataset of pharmaceutics disciplines is frequently challenging to comprehend. Since machine learning needs high-quality data sets, the open-source dataset can be a place to start. This work presents a systematic method to choose representative subsamples from the existing research, along with an extensive set of quality measures and a visualization strategy. The preceding article (Muthudoss et al.. in AAPS PharmSciTech 23, 2022) describes a workflow for leveraging near infrared (NIR) spectroscopy to obtain reliable and robust data on pharmaceutical samples. This study describes the systematic and structured procedure for selecting subsamples from the historical data. We offer a wide range of in-depth quality measures, diagnostic tools, and visualization techniques. A real-world, well-researched NIR dataset was employed to demonstrate this approach. This open-source tablet dataset ( http://www.models.life.ku.dk/Tablets ) consists of different doses in milligrams, different shapes, and sizes of dosage forms, slots in tablets, three different manufacturing scales (lab, pilot, production), coating differences (coated vs uncoated), etc. This sample is appropriate; that is, the model was developed on one scale (in this research, the lab scale), and it can be great to investigate how well the top models are transferable when tested on new data like pilot-scale or production (full) scale. A literature review indicated that the PLS regression models outperform artificial neural network-multilayer perceptron (ANN-MLP). This work demonstrates the selection of appropriate hyperparameters and their impact on ANN-MLP model performance. The hyperparameter tuning approaches and performance with available references are discussed for the data under investigation. Model extension from lab-scale to pilot-scale/production scale is demonstrated. HIGHLIGHTS: • We present a comprehensive quality metrics and visualization strategy in selecting subsamples from the existing studies • A comprehensive assessment and workflow are demonstrated using historical real-world near-infrared (NIR) data sets • Selection of appropriate hyperparameters and their impact on artificial neural network-multilayer perceptron (ANN-MLP) model performance • The choice of hyperparameter tuning approaches and performance with available references are discussed for the data under investigation • Model extension from lab-scale to pilot-scale successfully demonstrated.

Improvement of NIR prediction ability by dual model optimization in fusion of NSIA and SA methods

Feature selection and sample partitioning are both important to establish a quantitative analytical model for near-infrared (NIR) spectroscopy. The classical interval partial least squares (iPLS) model for waveband selection can be improved in combination of the simulated annealing (SA) algorithm. The sample set partitioning based on a joint x-y distance (SPXY) method for sample partitioning is based on the distances of both the x- and y- dimensions; it is expected to be optimized using the non-dominant sorting strategies (NS) combined with the immune algorithm (IA). In this study, we investigated the dual model optimization mode for simultaneous selection of feature waveband and sample partitioning, and proposed a novel method defined as SA-iPLS & SPXY-NSIA. The method explores a population evolution process, and takes the candidate individual as the link for the fusion optimization of SA-iPLS and SPXY-NSIA. The method screens feature wavebands and observes a good partition of the modeling samples, to construct a combined optimization strategy for fusion optimization of the target waveband and suitable sets of sample partitioning. The performance of the SA-iPLS & SPXY-NSIA method was tested using a soil sample dataset. To prove model enhancement, the proposed method was compared to the two traditional methods of Kennard-Stone (KS) and SPXY in combination with SA-iPLS. Experimental results show that the fusion model established by SA-iPLS & SPXY-NSIA performed better than the KS-SA-iPLS and SPXY-SA-iPLS models. The best testing results of the fusion model is with RMSET, RPD_T and R_T observed as 0.0107, 1.7233 and 0.9097, respectively. The proposed method is prospectively able to effectively improve the predictive ability of the NIR analytical model.

Multiblock variable influence on orthogonal projections (MB-VIOP) for enhanced interpretation of total, global, local and unique variations in OnPLS models

BACKGROUND

For multivariate data analysis involving only two input matrices (e.g., X and Y), the previously published methods for variable influence on projection (e.g., VIPOPLS or VIPO2PLS) are widely used for variable selection purposes, including (i) variable importance assessment, (ii) dimensionality reduction of big data and (iii) interpretation enhancement of PLS, OPLS and O2PLS models. For multiblock analysis, the OnPLS models find relationships among multiple data matrices (more than two blocks) by calculating latent variables; however, a method for improving the interpretation of these latent variables (model components) by assessing the importance of the input variables was not available up to now.

RESULTS

A method for variable selection in multiblock analysis, called multiblock variable influence on orthogonal projections (MB-VIOP) is explained in this paper. MB-VIOP is a model based variable selection method that uses the data matrices, the scores and the normalized loadings of an OnPLS model in order to sort the input variables of more than two data matrices according to their importance for both simplification and interpretation of the total multiblock model, and also of the unique, local and global model components separately. MB-VIOP has been tested using three datasets: a synthetic four-block dataset, a real three-block omics dataset related to plant sciences, and a real six-block dataset related to the food industry.

CONCLUSIONS

We provide evidence for the usefulness and reliability of MB-VIOP by means of three examples (one synthetic and two real-world cases). MB-VIOP assesses in a trustable and efficient way the importance of both isolated and ranges of variables in any type of data. MB-VIOP connects the input variables of different data matrices according to their relevance for the interpretation of each latent variable, yielding enhanced interpretability for each OnPLS model component. Besides, MB-VIOP can deal with strong overlapping of types of variation, as well as with many data blocks with very different dimensionality. The ability of MB-VIOP for generating dimensionality reduced models with high interpretability makes this method ideal for big data mining, multi-omics data integration and any study that requires exploration and interpretation of large streams of data.

Exploration of an odorous aldehydes and ketones produced by Uroglena americana using high resolution mass spectrometry, GC-Olfactometry, and multivariate analysis

Off-flavor events in tap water have been reported from various regions of Japan. Fishy smell is the second most common off-flavor in Japan and Uroglena americana (U. americana) is known to be a major contributor to the smell. However, the causative compound of the smell it produces still remains unrevealed to the best of our knowledge. In this study, an exploration of odorous aldehydes and ketones originating from U. americana was performed with a view to discovering a possible candidate substance of causative compounds. Environmental samples containing U. americana colony and cultured media with U. americana were analyzed with two high resolution mass spectrometers, one of them is coupled with liquid chromatography (LC-HRMS), and the other is with gas chromatography and a sniffing port (GC-O-HRMS). Multivariate analyses (MVA) were utilized to explore a compound that is likely to be odorous aldehydes or ketones with a reduced time of exploration. A combination of LC-HRMS and MVA resulted in the selection of one candidate substance and its formula was determined to be C₁₃H₂₀O₃ on the basis of its accurate mass and natural isotopic pattern. The candidate substance underwent GC-O-HRMS analyses and milk-like smell was detected at around its retention time. Although the detected smell was different from fishy smell, it is expected that the fishy smell is caused by multiple compounds to which the candidate substance belongs. First generation product ion spectra of the candidate substance suggested that it contains a hydroxyl group, a cyclohexene ring, and a ketone moiety.

Sample Consensus Model and Unsupervised Variable Consensus Model for Improving the Accuracy of a Calibration Model

In the quantitative analysis of spectral data, small sample size and high dimensionality of spectral variables often lead to poor accuracy of a calibration model. We proposed two methods, namely sample consensus and unsupervised variable consensus models, in order to solve the problem of poor accuracy. Three public near-infrared (NIR) or infrared (IR) spectroscopy data from corn, wine, and soil were used to build the partial least squares regression (PLSR) model. Then, Monte Carlo sampling and unsupervised variable clustering methods of a self-organizing map were coupled with the consensus modeling strategy to establish the multiple sub-models. Finally, sample consensus and unsupervised variable consensus models were obtained by assigning the weights to each PLSR sub-model. The calculated results show that both sample consensus and unsupervised variable consensus models can significantly improve the accuracy of the calibration model compared to the single PLSR model. The effectiveness of these two methods points out a new approach to achieve a further accurate result, which can take full advantage of the sample information and valid variable information.

Quantifying several adulterants of notoginseng powder by near-infrared spectroscopy and multivariate calibration

The authentication of traditional Chinese medicine (TCM) is critically important for public-health and economic terms. Notoginseng, a classical TCM of high economic and medical value, could be easily adulterated with Sophora flavescens powder (SFP), corn flour (CF) or other analogues of low-grade (ALG) because of their similar tastes, appearances and much lower cost. The main objective of this study was to evaluate the feasibility of applying of near-infrared (NIR) spectroscopy and multivariate calibration for identifying and quantifying several common adulterants in notoginseng powder. Two datasets were prepared for experiment. The competitive adaptive reweighted sampling (CARS) was used to select informative variables. Two different schemes were used for sample set partition. Model population analysis (MPA) was made. The results showed that, the constructed partial least squares (PLS) model using a reduced set of variables from CARS can provide superior performance to the full-spectrum PLS model. Also, the sample set partition is very of great importance. It seems that the combination of NIR spectroscopy, CARS and PLS is feasible to quantify common adulterants in notoginseng powder.

An efficient swarm intelligence approach to feature selection based on invasive weed optimization: Application to multivariate calibration and classification using spectroscopic data

Variable selection plays a key role in classification and multivariate calibration. Variable selection methods are aimed at choosing a set of variables, from a large pool of available predictors, relevant to the analyte concentrations estimation, or to achieve better classification results. Many variable selection techniques have now been introduced among which, those which are based on the methodologies of swarm intelligence optimization have been more respected during a few last decades since they are mainly inspired by nature. In this work, a simple and new variable selection algorithm is proposed according to the invasive weed optimization (IWO) concept. IWO is considered a bio-inspired metaheuristic mimicking the weeds ecological behavior in colonizing as well as finding an appropriate place for growth and reproduction; it has been shown to be very adaptive and powerful to environmental changes. In this paper, the first application of IWO, as a very simple and powerful method, to variable selection is reported using different experimental datasets including FTIR and NIR data, so as to undertake classification and multivariate calibration tasks. Accordingly, invasive weed optimization - linear discrimination analysis (IWO-LDA) and invasive weed optimization- partial least squares (IWO-PLS) are introduced for multivariate classification and calibration, respectively.

Variable selection based on clustering analysis for improvement of polyphenols prediction in green tea using synchronous fluorescence spectra

Synchronous fluorescence spectra, combined with multivariate analysis were used to predict flavonoids content in green tea rapidly and nondestructively. This paper presented a new and efficient spectral intervals selection method called clustering based partial least square (CL-PLS), which selected informative wavelengths by combining clustering concept and partial least square (PLS) methods to improve models' performance by synchronous fluorescence spectra. The fluorescence spectra of tea samples were obtained and k-means and kohonen-self organizing map clustering algorithms were carried out to cluster full spectra into several clusters, and sub-PLS regression model was developed on each cluster. Finally, CL-PLS models consisting of gradually selected clusters were built. Correlation coefficient (R) was used to evaluate the effect on prediction performance of PLS models. In addition, variable influence on projection partial least square (VIP-PLS), selectivity ratio partial least square (SR-PLS), interval partial least square (iPLS) models and full spectra PLS model were investigated and the results were compared. The results showed that CL-PLS presented the best result for flavonoids prediction using synchronous fluorescence spectra.

Detection and quantification of food colorant adulteration in saffron sample using chemometric analysis of FT-IR spectra

The aim of present study is to investigate the combination of Fourier transform infrared spectroscopy with pattern recognition to recognize the standard saffron from those which have been adulterated with various types of food colorants.