Review
Instruments & Instrumentation
J. Renwick Beattie, Francis W. L. Esmonde-White
Summary: Spectroscopy rapidly captures a large amount of data, which is processed using principal component analysis to simplify complex spectral datasets into comprehensible information. Despite its wide use, the linear algebra behind principal component analysis is often not well understood by applied scientists and spectroscopists. The process traces the journey of spectra and relies solely on the information within the spectra to provide meaningful interpretation and analysis.
APPLIED SPECTROSCOPY
(2021)
Article
Chemistry, Analytical
Raffaele Vitale, Marina Cocchi, Alessandra Biancolillo, Cyril Ruckebusch, Federico Marini
Summary: This article provides a comprehensive tutorial on the classification method SIMCA, offering pragmatic guidelines for its correct utilization and answering basic questions on why, when, and how to employ SIMCA. The article addresses mathematical and statistical fundamentals, describes different variants of the algorithm, provides a flowchart for model parameter tuning, illustrates model assessment tools, and gives computational details and suggestions for model validation.
ANALYTICA CHIMICA ACTA
(2023)
Article
Computer Science, Interdisciplinary Applications
Yushan Liu, Luyi Li, Zeming Chang, Pan Wang
Summary: This paper proposes a Kriging-based analytical technique for estimating multivariate sensitivity indices (MSI). Two types of MSI are studied and analyzed, one based on principal component analysis (MSI-PCA) and the other based on covariance decomposition (MSI-CD). The accuracy and efficiency of the proposed KBA and D-Kriging methods are tested and discussed using four examples.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2022)
Article
Biochemistry & Molecular Biology
Abdulrahman Aljannahi, Roudha Abdulla Alblooshi, Rashed Humaid Alremeithi, Ioannis Karamitsos, Noora Abdulkarim Ahli, Asma Mohammed Askar, Ikhlass Mohammed Albastaki, Mohamed Mahmood Ahli, Sanjay Modak
Summary: Synthetic fibers can provide valuable evidence in crime scenes and help establish connections between suspects, victims, and crime scenes. This study used Fourier transform infrared spectroscopy and multivariate statistical methods combined with machine learning classification models to classify synthetic textile fibers successfully.
Article
Agricultural Economics & Policy
Salvatore Ciano, Lucia Maddaloni, Mattia Rapa, Anna Maria Tarola
Summary: This paper proposes a multi-methodological chemical profiling and multivariate data analysis of organic hempseed oil samples from the retail market. The result of the analysis confirmed the ideal fatty acid ratio and excellent content of antioxidant species in the hempseed oil. The study also utilized discriminant analyses to classify samples according to label information, achieving high classification efficiency for extraction procedure verification, geographical origin, and prices.
BRITISH FOOD JOURNAL
(2023)
Article
Medicine, Legal
Anais Hermelin, Loic Fabien, Julia Fischer, Nikola Saric, Genevieve Massonnet, Celine Burnier
Summary: The study investigated vaginal matrix residues using DRIFTS-FTIR and py-GC/MS, identifying proteins and lipids in the samples but no silicone residues. This has promising implications for forensic evidence interpretation, with further research needed to validate models and assess limitations in casework conditions.
FORENSIC SCIENCE INTERNATIONAL
(2021)
Article
Food Science & Technology
R. Rios-Reina, R. M. Callejon, J. M. Amigo
Summary: This study used near-infrared hyperspectral imaging for the spectral study of Spanish and Chinese pine nuts, identifying chemical distribution and composition. A classification model was developed for quality control with high accuracy.
Article
Agriculture, Dairy & Animal Science
Mohamed M. Salem, Mohammed A. F. Nasr, Amin M. S. Amin
Summary: This study estimated genetic parameters and breeding values for birth weight and reproductive and milk traits of Egyptian buffalo. The study found that birth weight and reproduction traits would respond slowly to selection, while production traits would respond faster. The results suggest that principal component analysis could be used in genetic improvement programs for Egyptian buffalo instead of traditional selection indices.
TROPICAL ANIMAL HEALTH AND PRODUCTION
(2021)
Article
Chemistry, Applied
Guido Rolandelli, M. Pilar Buera, Silvio D. Rodriguez
Summary: This study investigates the effectiveness of Fourier Transform Infrared spectroscopy (FTIR) combined with chemometrics in rapidly and non-destructively monitoring the thermal stability of pure sesame oil (SeO) as well as SeO blends adulterated with corn, soybean, and sunflower oils. The results show that FTIR provides a sensitive approach for analyzing thermal degradation and compositional changes of pure and adulterated SeO under controlled temperature conditions. Principal component analysis (PCA) can distinguish pure SeO at different treatment conditions, while soft independent modeling class analogy (SIMCA) can differentiate between pure and adulterated samples with high precision and accuracy. Finally, partial least squares regression (PLSR) models exhibit a good linear correlation between FTIR signals at different conditions and the proportion of pure SeO.
JOURNAL OF FOOD COMPOSITION AND ANALYSIS
(2024)
Article
Instruments & Instrumentation
Mario Marchetti, Jean-Michel Mechling, Sarah Janvier-Badosa, Marc Offroy
Summary: This paper presents an original method for monitoring the hydration reaction of cement using non-invasive Raman spectroscopy, coupled with chemometrics analysis. The results are consistent with thermogravimetric analysis and accurately reveal the different phases involved in the hydration and curing process.
APPLIED SPECTROSCOPY
(2023)
Article
Management
Arthur Charpentier, Stephane Mussard, Tea Ouraga
Summary: The Gini PCA, based on the generalized Gini correlation index, is proposed as a method for robust dimensionality reduction that is shown to be equivalent to standard PCA in the Gaussian case. Monte Carlo simulations and application on cars data demonstrate the robustness and different interpretations of results compared to variance PCA.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2021)
Article
Biochemical Research Methods
D. M. Mazur, A. A. Sosnova, T. B. Latkin, B. Artaev, K. Siek, D. A. Koluntaev, A. T. Lebedev
Summary: Thorough non-target analysis of snow samples from Arkhangelsk city revealed the presence of hundreds of organic compounds, including some priority pollutants from the US EPA list. However, the levels of these compounds were found to be much lower than the safe thresholds established in Russia. Phenol and dioctylphthalate were identified as pollutants of concern due to their relatively higher levels. Car traffic and pulp and paper mills were proposed as possible sources of the major pollutants. Various tools and software protocols were applied to process the data and the clusterization results showed similarities with certain peculiarities.
ANALYTICAL AND BIOANALYTICAL CHEMISTRY
(2023)
Article
Materials Science, Characterization & Testing
Ivana Chamradova, Pavel Porizka, Jozef Kaiser
Summary: This paper analyzes polymer materials using LIBS, studying them in different atmospheres and examining their signal responses. It finds that the emission spectra of various polymers are similar and focuses on the sources of molecular bands.
Article
Chemistry, Applied
Carla de Fatima Alves Nonato, Cicera Janaine Camilo, Debora Odilia Duarte Leite, Mario Gustavo Lucio Albuquerque da Abrega, Jaime Ribeiro-Filho, Irwin Rose Alencar de Menezes, Josean Fechine Tavares, Jose Galberto Martins da Costa
Summary: This study compared the chemical composition and antioxidant potential of essential oils from the Lippia genus through chemometric analysis. The results showed that the essential oils exhibited significant antioxidant activity, with different compounds in Lippia alba being effective in various antioxidant methods. The findings highlight the importance of Lippia compounds for protecting against oxidative stress and their potential use in food preservation.
Article
Food Science & Technology
J. C. Castura, P. Varela, T. Naes
Summary: Principal component analysis (PCA) is commonly used to summarize and explore multivariate data sets, including sensory evaluation data sets. This study proposes a method to conduct PCA on a results matrix where only a subset of paired comparisons is of interest. The proposed approach is illustrated with two data sets from trained sensory panels. The results show that the PCA conducted with the proposed method extracts more variance from the relevant paired comparisons and better separates the relevant pairs compared to conventional PCA.
FOOD QUALITY AND PREFERENCE
(2023)
Article
Chemistry, Analytical
Consuelo Pizarro, Irene Arenzana-Ramila, Nuria Perez-del-Notario, Patricia Perez-Matute, Jose Maria Gonzalez-Saiz
ANALYTICA CHIMICA ACTA
(2016)
Article
Food Science & Technology
Claudia A. Teixeira dos Santos, Ricardo N. M. J. Pascoa, Mafalda Cruz Sarraguca, Patricia A. L. S. Porto, Antonio L. Cerdeira, J. M. Gonzalez-Saiz, C. Pizarro, Joao A. Lopes
FOOD RESEARCH INTERNATIONAL
(2017)
Article
Biochemical Research Methods
Consuelo Pizarro, Isabel Esteban-Diez, Irene Arenzana-Ramila, Jose M. Gonzalez-Saiz
JOURNAL OF BIOPHOTONICS
(2018)
Article
Chemistry, Analytical
Claudia A. Teixeira dos Santos, Ricardo N. M. J. Pascoa, Patricia A. L. S. Porto, Antonio L. Cerdeira, J. M. Gonzalez-Saiz, C. Pizarro, Joao A. Lopes
Article
Chemistry, Applied
C. Pizarro, S. Rodriguez-Tecedor, I. Esteban-Diez, N. Perez-del-Notario, J. M. Gonzalez-Saiz
Article
Chemistry, Applied
J. M. Gonzalez-Saiz, I. Esteban-Diez, S. Rodriguez-Tecedor, N. Perez-del-Notario, I. Arenzana-Ramila, C. Pizarro
Article
Chemistry, Analytical
C. Pizarro, N. Perez-del-Notariol, A. Saenz-Mateo, J. M. Gonzalez-Saiz
Article
Chemistry, Analytical
Kateryna Tkachenko, Maria Espinosa, Isabel Esteban-Diez, Jose M. Gonzalez-Saiz, Consuelo Pizarro
Summary: An untargeted FTIR metabolomic approach was used to study metabolic changes in Parkinson's disease. A classification strategy based on SELECT-LDA was proposed, achieving high correct assignment rates in distinguishing PD patients from AD patients and healthy controls, as well as stratifying different stages of PD and differentiating between PDD and AD. The selected metabolic signatures could be used for screening and diagnosis.
Article
Chemistry, Analytical
Consuelo Pizarro, Isabel Esteban-Diez, Maria Espinosa, Fernando Rodriguez-Royo, Jose-Maria Gonzalez-Saiz