An extension of PPLS-DA for classification and comparison to ordinary PLS-DA

A Comparison of Lipid Contents in Different Types of Peanut Cultivars Using UPLC-Q-TOF-MS-Based Lipidomic Study

Peanuts are a rich dietary source of lipids, which are essential for human health. In this study, the lipid contents of 13 peanut cultivars were analyzed using UPLC-Q-TOF-MS and GC–MS. The OXITEST reactor was used to test their lipid oxidation stabilities. A total of 27 subclasses, 229 individual lipids were detected. The combined analysis of lipid and oxidation stability showed that lipid unsaturation was inversely correlated with oxidation stability. Moreover, lipid profiles differed significantly among the different peanut cultivars. A total of 11 lipid molecules (TG 18:2/18:2/18:2, TG 24:0/18:2/18:3, TG 20:5/14:1/18:2, TG 18:2/14:1/18:2, PE 17:0/18:2, BisMePA 18:2/18:2, PG 38:5, PMe 18:1/18:1, PC 18:1/18:1, MGDG 18:1/18:1, TG 10:0/10:1/18:1) might be employed as possible indicators to identify high oleic acid (OA) and non-high OA peanut cultivars, based on the PLS-DA result of lipid molecules with a VIP value greater than 2. This comprehensive analysis will help in the rational selection and application of peanut cultivars.

PLS Generalized Linear Regression and Kernel Multilogit Algorithm (KMA) for Microarray Data Classification Problem

This study involves the implentation of the extensions of the partial least squares generalized linear regression (PLSGLR) by combining  it with logistic regression and  linear  discriminant analysis,  to  get a  partial least  squares generalized linear  regression-logistic regression model (PLSGLR-log),  and a partial least squares generalized linear regression-linear discriminant analysis model (PLSGLRDA). A comparative  study  of  the obtained  classifiers with   the   classical  methodologies like  the k-nearest  neighbours (KNN), linear   discriminant  analysis  (LDA),   partial  least  squares discriminant analysis (PLSDA),  ridge  partial least squares (RPLS), and  support vector machines(SVM)  is  then  carried  out.    Furthermore,  a  new  methodology known as kernel multilogit algorithm (KMA) is also implemented and its performance compared with those of the other classifiers. The KMA emerged as the best classifier based  on the lowest  classification error  rates  compared to  the  others  when  applied   to  the  types   of data   are considered;  the  un- preprocessed and preprocessed.

Comparing the qualitative performances of handheld NIR and Raman spectrophotometers for the detection of falsified pharmaceutical products

Over the last decade, the growth of the global pharmaceutical market has led to an overall increase of substandard and falsified drugs especially on the African market (or emerging countries). Recently, several methods using handheld/portable vibrational spectroscopy have been developed for rapid and on-field drug analysis. The objective of this work was to evaluate the performances of various NIR and Raman handheld spectrophotometers in specific brand identification of medicines through their primary packaging. Three groups of drug samples (artemether-lumefantrine, paracetamol and ibuprofen) were used in tablet or capsule forms. In order to perform a critical comparison, the analytical performances of the two analytical systems were compared statistically using three methods: hierarchical clustering algorithm (HCA), data-driven soft independent modelling of class analogy (DD-SIMCA) and hit quality index (HQI). The overall results show good detection abilities for NIR systems compared to Raman systems based on Matthews's correlation coefficients, generally close to one. Raman systems are less sensitive to the physical state of the samples than the NIR systems, it also suffers of the auto-fluorescence phenomenon and the signal of highly dosed active pharmaceutical ingredient (e.g. paracetamol or lumefantrine) may mask the signal of low-dosed and weaker Raman active compounds (e.g. artemether). Hence, Raman systems are less effective for specific product identification purposes but are interesting in the context of falsification because they allow a visual interpretation of the spectral signature (presence or absence of API).

A new method combining LDA and PLS for dimension reduction

Linear discriminant analysis (LDA) is a classical statistical approach for dimensionality reduction and classification. In many cases, the projection direction of the classical and extended LDA methods is not considered optimal for special applications. Herein we combine the Partial Least Squares (PLS) method with LDA algorithm, and then propose two improved methods, named LDA-PLS and ex-LDA-PLS, respectively. The LDA-PLS amends the projection direction of LDA by using the information of PLS, while ex-LDA-PLS is an extension of LDA-PLS by combining the result of LDA-PLS and LDA, making the result closer to the optimal direction by an adjusting parameter. Comparative studies are provided between the proposed methods and other traditional dimension reduction methods such as Principal component analysis (PCA), LDA and PLS-LDA on two data sets. Experimental results show that the proposed method can achieve better classification performance.