Journal of Chemometrics最新文献

Hyphenated chromatographic techniques are widely used to analyze and characterize complex samples. Chemometric methods are generally needed to extract the qualitative and quantitative information of the target analytes from complex hyphenated chromatographic data. However, neither two-way nor three-way chemometric methods are efficient enough in analyzing hyphenated chromatographic data with both severe peak overlapping and retention time shift across samples. To address this issue, a practical framework was proposed herein. It consists of three chemometric algorithms, that is, (1) “fix-sized moving window evolving target spectral projection” for locating the possible peak positions of the target analytes, (2) “target identification based on singular value comparison” for determining whether the identified peaks are indeed the chromatographic peaks of the target analytes, and (3) “fix-sized moving window evolving trilinear decomposition” for obtaining the quantitative results of the target analytes. Experimental results on the GC-MS data sets of mixture samples of 10 compounds verified that the proposed framework could deal with the problems of both severe peak overlapping and retention time shift across samples. The proposed framework has the advantages of simplicity in concept, easy implementation, and good performance and hence is expected to be a competitive alternative to existing methods for the analysis of hyphenated chromatographic data of complex samples.

Imprinted proteins (IPs) are promising alternatives to natural recognition systems, such as biological receptors or antibodies. One of the crucial stages during development of IPs is removal of the template molecules from its complex with the protein. In this study, bovine serum albumin was imprinted in the presence of 4-hydroxycoumarin (4-HC); purification of IPs were carried out by dialysis, and fluorescence 3D spectroscopy was used to monitor the IP purification process. Excitation–emission matrix (EEM) was further investigated via several chemometric algorithms (principal component analysis [PCA], parallel factor analysis [PARAFAC], and independent components analysis [ICA]). We found that the models using PARAFAC and ICA worked better than those of PCA. It was shown that PARAFAC and ICA analyses allow not only to recognize IP sample with signal close to nonimprinted protein, but also to provide recommendations on the optimal dialysis time.

Characterizing complicated solution phase systems in situ requires advanced modeling techniques to capture the intricate balances between the many chemical species. Due to the error inherent in any scientific measurement, a spectrophotometric titration experiment with nickel(II) and ethylenediamine (en) was repeated six times using an autotitrator to test the replicability of the data and the consistency of the resulting thermodynamic model. All six datasets could be modeled very tightly (R² > 99.9999%) with the following eight complexes: [Ni]²⁺, [Ni₂en]⁴⁺, [Nien]²⁺, [Ni₂en₃]⁴⁺, [Nien₂]²⁺, [Ni₂en₅]⁴⁺, [Nien₃]²⁺, and [Nien₆]²⁺. The logK values for the stepwise associative reactions agree with existing literature values for the majority species ([Nien_{n = 1–3}]²⁺) and matched expectations for the minority species; 95% confidence intervals for each logK value were determined via bootstrapping, which quantifies the variability in the binding constant value that is supported by a given dataset. The repeated experiments, which could not be successfully concatenated together, demonstrate that replication is crucial to capturing all the variability in the logK values. Conversely, bootstrapped confidence intervals across multiple experiments can be readily combined to generate an appropriate range for an experimentally determined binding constant.

An overview of the status of the research in analytical figures of merit is provided, including all calibration scenarios from univariate to multivariate and multiway analytical protocols. Both linear and nonlinear multivariate models are considered. Starting with the simplest multivariate model, inverse least-squares regression, the basic concepts of sensitivity, sample leverage, and limit of detection are introduced. The extension to other multivariate models is discussed, as well as to nonlinear models based on radial basis functions, kernel partial least-squares, and multilayer feed-forward artificial neural networks. Finally, multiway calibration models are discussed, including multilinear decomposition models such as parallel factor analysis (PARAFAC) and multivariate curve resolution–alternating least-squares (MCR-ALS). In the latter case, recent developments concerning the pervasive phenomenon of rotational ambiguity are discussed. Unfinished works and areas where further research efforts are needed to develop closed-form expressions and to fully understand their meaning are included.

Deep learning has attracted widespread attention in data modeling and key quality indicator prediction in the chemical industry. However, traditional deep learning networks usually distort the original data distribution due to the superposition effect of multiple layers of nonlinear activation functions. In this case, multivariate statistical learning techniques present an avenue to reveal the intrinsic relationship of the data by combining the linear trends between input and predictor variables. To comprehensively capture data features from multiple perspectives, this study proposes a deep learning-based data modeling network called the information retention unit (IRU). This network combines intrinsic attributes to partial least squares (PLS) and autoencoder (AE) modalities, thus engendering an adaptive response to the complex linear and nonlinear data features. Furthermore, multiple IRUs can be stacked to construct a deep information retention network (DIRN), which enhances the robust extraction of deep data features. Finally, the effectiveness of the proposed network is validated through its prediction application on a dataset obtained from a real-world chemical industrial process. This method combines multivariate statistical learning techniques based on deep learning, providing an innovative and practical solution for data analysis and prediction in the chemical industry.

The rapidly changing climatic scenarios are highly favorable for the rising diseases that lead to increasing threats to food production and supply. Various scholars and scientists make long steps to hasten the process of making innovations in farming for managing these issues. In this context, UAV is applied for the purpose of managing and monitoring plant health. The abiotic stresses available in plant diagnosis through traditional strategies are highly labor-intensive and unfit for large-scale deployment. Conversely, UAVs designed with mobile sensors, multispectral, radar, and so on make them flexible, affordable, and more effective. Thus, this study proposes a novel meta ensemble transfer extreme gradient-based random tactical unit (MTEG-RTU) algorithm for diagnosing crop illnesses precisely. The proposed MTEG-RTU methodology entails three methods such as transfer learning, adaptive boost, and meta-ensemble, and the hyper parameters are tuned using random tactical unit algorithm. Healthier and disordered crop images gained from the crop disease dataset comprise 8000 images and are preprocessed. The more optimal features from the preprocessed images are learned through the ResNet method, and these features enter into the classification phase. Random tactical unit algorithm enhanced the performance by optimizing the hyperparameters of MTEG classifier. The experimental results conducted based on the various assessment components and validation dataset indicate that the developed method outperformed the other chosen models, achieving precision, recall, and accuracy of 98.5%, 97.9%, and 98.6%, respectively. The other achievements made by the model are offering technical guidance for conducting the precise diagnosis and treatment of plant pathologies with less time of 9 s.