Partial Least Squares Research Articles

Comparison Between the Partial Least Squares Method and Principal Components Using Genetic Algorithm with an Application

The problem of multicollinearity among independent variables in a regression model was addressed in this study using Partial Least Squares (PLS) and Principal Component Analysis (PCA). The influence exerted by multicollinearity can deteriorate the results of regression modeling; that is, it makes the traditional technique OLS less reliable. The research is centered on applying advanced algorithms for Partial Least Squares (SIMPL and O-PLS) as well as for PCA (NIPALS and SVD) on real-life data scenarios, as well as integrating genetic algorithm (GAs) with these algorithms to optimize predictive performance. The relative efficiency of these methods is evaluated primarily through the amplitude of the Mean Square Error (MSE) used as a criterion for comparison. The results show the effectiveness of PLS-OPLS above PCA in terms of the lowest before and after embedding genetic algorithms into the MSE. All this underpins the effectiveness of PLS in minimizing multicollinearity thereby allowing for the formulation and prediction of very highly predictive models. More and more indication of subtle class was revealed with regard to fine tuning advanced GA technique in favor of enhancing regression modeling for complex data analysis. The ongoing research will help in opening numerous possibilities to reinforce regression methodology, especially when one factor in an array of applications representing a relatively significant level of prediction accuracy.

Machine learning driven benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality.

The aim of this study was to explore application of visible and near-infrared (Vis/NIR) spectroscopy combined with machine learning models for SSC and TA prediction of hybrid citrus. The Vis/NIR spectra of samples including navel-region, equator-region and multi-region combination spectra in navel-region and equator-region were collected using a benchtop equipment. The performance of SSC and TA prediction models with different region spectra, including partial least squares (PLS), random forest (RF), k-nearest neighbors (KNN), support vector machine (SVM) and multilayer feedforward neural network (MFNN), was assessed. The accuracy of SSC and TA prediction models with multi-region combination (raw) spectra was better compared to navel-region and equator-region, suggesting that multi-region combination spectra collection method was more suitable. Subsequently, the spectral pre-processing, including Savitzky-Golay smoothing (SGS), maximum normalization (MN), multiplicative scatter correction (MSC), linear baseline correction (LBC) and first derivative (1stD), were performed. The performance of SSC and TA prediction models with different pre-processing spectra was further compared. The PLS with SGS spectra (SGS-PLS) and MFNN with raw spectra (Raw-MFNN) exhibited superior validation effects for SSC and TA prediction, respectively. In a subsequent prediction in new samples, SGS-PLS achieved an RP2 of 0.875, an RMSEP of 0.572% and a MAEP of 0.469% for SSC prediction, and Raw-MFNN achieved an RP2 of 0.800, an RMSEP of 0.0322% and a MAEP of 0.0249% for TA prediction, indicating excellent generalization ability. These results indicate the great potential of benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality at mass-scale level.

An exploratory study on the development of a sensory wheel affiliated with the emotional lexicon for chrysanthemum infusion.

Chrysanthemum infusion has gained popularity outside of the East Asian market in recent years. However, despite the growing international demand, standardized methods to evaluate the sensory attributes of chrysanthemum tea, which are essential for guiding consumers and ensuring quality control, remain underdeveloped. In this study, a trained panel conducted a quantitative descriptive analysis (QDA®) of eight chrysanthemum samples, successfully distinguishing between them. Additionally, consumer emotional responses to chrysanthemum infusion were assessed using the Check-All-That-Apply (CATA) method. By mapping the descriptive sensory data with emotional response data through Partial Least Squares Regression (PLSR), we identified sensory drivers that elicit specific emotional responses. For this analysis, the emotion lexicon from EsSense25 was clustered into six dimensions: pleasant, tame, warm, active, negative, and bored. Our findings indicate that attributes such as smoothness and chrysanthemum-cucumber flavour induce calm and reassuring emotions (e.g., secure, understanding, calm, and tame). Conversely, sweetness and floral flavour are associated with positive emotions such as happiness, joy, and general well-being, suggesting that floral sweet beverages can mitigate negative emotions. However, sensory attributes such as bitterness, astringency, and vegetal aroma were linked to negative emotions and were sometimes associated with feelings of activity. The development of a sensory wheel, integrated with the emotional lexicon for chrysanthemum infusion, provides a tool for identifying sensory drivers behind emotional experiences. This tool offers valuable insights for market applications and product development, enhancing consumer satisfaction by aligning product attributes with desired emotional outcomes.

Spatiotemporal evolution and driving forces of landscape structure and habitat quality in river corridors with ceased flow: A case study of the Yongding River corridor in Beijing, China.

Flow cessation leads to severe degradation of river corridor landscape structure, habitat quality, and ecological functions. This study focuses on the representative river with ceased flow in northern China, the Yongding River plain section. Utilizing long-term, high-resolution satellite remote sensing imagery and the InVEST model, we analyzed the spatiotemporal evolution of landscape structure and habitat quality (HQ) before and after river corridor flow cessation over the past 50 years. The study further employs partial least squares regression (PLSR) to explore the impact of landscape structural changes on HQ and uses generalized additive models (GAMs) and geographical detector (GeoDetector) to quantitatively identify key factors affecting habitat degradation and their interactive effects. Results indicate that from 1967 to 2018, mid-channel bar, floodplain, and waterbody decreased sharply from 37.4% to 3.8%. The mean HQ value dropped from 0.58 to 0.34 after flow cessation. Although HQ slightly recovered post-2004, high-quality habitat areas remain absent. Different landscape structures significantly influence HQ, with increased size and area of the waterbody and forest patches positively contributing, while cultivated land, barren land, and built-up land generally have negative impacts. PLAND, LPI, MPS, and AWMPFD are key metrics for optimizing landscape structure and implementing habitat restoration in river management. Anthropogenic activities emerged as the primary driver of river corridor habitat degradation post-flow cessation. Different drivers exhibit complex linear and nonlinear effects on HQ. Based on these findings, we propose ecological management strategies for river corridors with ceased flow. This study is essential for a deeper understanding of river corridors' structural dynamics and degradation mechanisms, providing a scientific basis for effective ecological restoration and management.

Building a near-infrared (NIR) soil spectral dataset and predictive machine learning models using a handheld NIR spectrophotometer.

This near-infrared spectral dataset consists of 2,106 diverse mineral soil samples scanned, on average, on six different units of the same low-cost commercially available handheld spectrophotometer. Most soil samples were selected from the USDA NRCS National Soil Survey Center-Kellogg Soil Survey Laboratory (NSSC-KSSL) soil archives to represent the diversity of mineral soils (0-30 cm) found in the United States, while 90 samples were selected from Ghana, Kenya, and Nigeria to represent available African soils in the same archive. All scanning was performed on dried and sieved (<2 mm) soil samples. Machine learning predictive models were developed for soil organic carbon (SOC), pH, bulk density (BD), carbonate (CaCO3), exchangeable potassium (Ex. K), sand, silt, and clay content from their spectra in the R programming language using most of this dataset (1,976 US soils) and are included in this data release. Two model types, Cubist and partial least squares regression (PLSR) were developed using two strategies: (1) using an average of the spectral scans across devices for each sample and, (2) using the replicate spectral scans across devices for each sample. We present the internal performance of these models here. The dry spectra and Cubist models for these soil properties are available for download from 10.5281/zenodo.7586621. An example of detailed code used to produce these models is hosted at the Open Soil Spectral Library, a free service of the Soil Spectroscopy for the Global Good Network (soilspectroscopy.org), enabling broad use of these data for multiple soil monitoring applications.

Quantitative Analysis of Peanut Skin Adulterants by Fourier Transform Near-Infrared Spectroscopy Combined with Chemometrics

Peanut skin is a potential medicinal material. The adulteration of peanut skin samples with starchy substances severely affects their medicinal value. This study aimed to quantitatively analyze the adulterants present in peanut skin using Fourier transform near-infrared (FT-NIR) spectroscopy. Two adulterants, sweet potato starch and corn starch, were included in this study. First, spectral information of the adulterated samples was collected for characterization. Then, the applicability of different preprocessing methods and techniques to the obtained spectral data was compared. Subsequently, the Competitive Adaptive Reweighted Sampling (CARS) algorithm was used to extract effective variables from the preprocessed spectral data, and Partial Least Squares Regression (PLSR), a Support Vector Machine (SVM), and a Black Kite Algorithm-Support Vector Machine (BKA-SVM) were employed to predict the adulterant content in the samples, as well as the overall adulteration level. The results showed that the BKA-SVM model performed excellently in predicting the content of sweet potato starch, corn starch, and overall adulterants, with determination coefficients (RP2) of 0.9833, 0.9893, and 0.9987, respectively. The experimental results indicate that FT-NIR spectroscopy combined with advanced machine learning techniques can effectively and accurately detect adulterants in peanut skin, providing a reliable technological support for food safety detection.

Cell Culture Media and Raman Spectra Preprocessing Procedures Impact Glucose Chemometrics

ABSTRACTDeployment of process analytical technology tools such as Raman or IR spectroscopy and associated multivariate calibration models for process monitoring and control plays an important role in process automation and advanced manufacturing of pharmaceuticals. Preprocessing or preparation of the spectroscopic data is an important step in developing a multivariate calibration model. There are several ways available to preprocess the data and each may influence the calibration model performance differently. Here we investigated the influence of preprocessing procedures on the development and performance of the chemometric models to predict the glucose concentration in a bioreactor. Box–Behnken design of experiment (DOE) was used to generate the Raman spectroscopy data. Four factors were considered critical in the DOE—glucose, glutamine, glutamic acid, and antifoam concentration. Raman spectroscopy data were collected both with and without aeration conditions, independently from three cell culture media. For each medium, data consisted of calibration set (27 conditions) and model validation set (9 conditions) separately. Additionally, Raman data was also collected for certain DOE runs with increasing concentration of cell densities ranging from 0.5 × 10 E06/mL to 30 × 10 E06/mL under aerating conditions. Data from the three cell culture media were used separately to develop calibration models that used four different preprocessing procedures, namely, baseline correction (BLC), Savitzky–Golay smoothing (SGS), Savitzky–Golay derivative (SGD) and orthogonal signal correction (OSC). The preprocessing procedures were applied individually and in combinations to evaluate the calibration model parameters and the performance metrics. We further developed glucose calibration models based on partial least squares (PLS) regression with 1–3 principal components. The models developed with OSC procedure gave superior performance metrics with just one principal component across all three media. Models developed with other preprocessing procedures required two or more principal components to give comparable performance. Overall, the choice of preprocessing procedures affected the model performance.

Unravelling critical climatic factors and phenological stages impacting spring barley yields across Europe

Unravelling critical climatic factors and phenological stages impacting spring barley yields across Europe

Intelligent consensus-based predictions of early life stage toxicity in fish tested in compliance with OECD Test Guideline 210.

Early life stage (ELS) toxicity testing in fish is a crucial test procedure used to evaluate the long-term effects of a wide range of chemicals, including pesticides, industrial chemicals, pharmaceuticals, and food additives. This test is particularly important for screening and prioritizing thousands of chemicals under the Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH) regulation. In silico methods can be used to estimate the toxicity of a chemical when no experimental data is available and to reduce the cost, time, and resources involved in the experimentation process. In the present study, we developed predictive Quantitative Structure-Activity Relationship (QSAR) models to assess chronic effects of chemicals on ELS in fish. Toxicity data for ELS in fish was collected from two different sources, i.e. J-CHECK and eChemPortal, which contain robust study summaries of experimental studies performed according to OECD Test Guideline 210. The collected data included two types of endpoints - the No Observed Effect Concentration (NOEC) and the Lowest Observed Effect Concentration (LOEC), which were utilized to develop the QSAR models. Six different partial least squares (PLS) models with various descriptor combinations were created for both endpoints. These models were then employed for intelligent consensus-based prediction to enhance predictability for unknown chemicals. Among these models, the consensus model - 3 (Q2F1 = 0.71, Q2F2 = 0.71) and individual model - 3 (Q2F1 = 0.80, Q2F2 = 0.79) exhibited most promising results for both the NOEC and LOEC endpoints. Furthermore, these models were validated experimentally using experimental data from nine different industrial chemicals provided by Global Product Compliance (Europe) AB. Lastly, the models were used to screen and prioritize chemicals obtained from the Pesticide Properties (PPDB) and DrugBank databases.

A generalizable methodology for predicting retention time of small molecule pharmaceutical compounds across reversed-phase HPLC columns.

Quantitative structure retention relation (QSRR) is an active field of research, primarily focused on predicting chromatography retention time (Rt) based on molecular structures of an input analyte on a single or limited number of reversed-phase HPLC (RP-HPLC) columns. However, in the pharmaceutical chemistry manufacturing and controls (CMC) settings, single-column QSRR models are often insufficient. It is important to translate retention time across different HPLC methods, specifically different stationary phases (SP) and mobile phases (MP), to guide the HPLC method development, and to bridge organic impurity profiles across different development phases and laboratories. In response to this need, we present a novel approach for retention time transfer across SPs and MPs, without requiring pre-existing Rt data on the target column. To achieve this, we developed an RP-HPLC based Genentech Multi-column Retention Time (GMCRT) database containing 51 small molecule pharmaceutical compounds analyzed on twenty SPs and multiple pH levels. The database incorporated the SP selectivity parameters from Hydrophobic Subtraction Model (HSM) - hydrophobicity (H), steric hindrance (S), hydrogen-bond acidity (A), hydrogen-bond basicity (B), ionic interaction (C) under two different pHs (2.8 and 7) and ethylbenzene (EB) retention factor. Two machine learning approaches, partial least squares (PLS) and artificial neural networks (ANN) were found to improve accuracy of Rt prediction on new SPs compared to the direct mapping approach that have been previously published, especially when the RP-HPLC columns have significant selectivity difference. As a comparison, our approach does not require pre-existing retention data on the target SPs and it is generalizable to any RP-HPLC columns with a set of known column selectivity parameters (https://www.hplccolumns.org/). The generalizability is achievable not only via the available retention data correlation among the twenty commonly-used RP-HPLC columns in GMCRT, but also via the retrainable mechanism of our ML models by adding Rt of the compounds of interest on the source columns into GMCRT, followed by predicting Rt on the target column. Thus, we propose a new QSRR framework that incorporates the physiochemical properties of SPs and MPs and makes the retention time prediction transferable across SPs and MPs. Such a framework is expected to open up possibilities for developing more comprehensive and generalizable models, and streamline RP-HPLC method development and lifecycle management across various pharmaceutical CMC development phases.

Robust near-infrared modeling for pharmaceutical powder streams: External variable augmented iterative optimization technology (EVA-IOT).

The adoption of pure component models, such as iterative optimization technology (IOT) algorithms, is gaining significant interest in the pharmaceutical industry, primarily because of their calibration-free/minimal calibration requirements for process analytical technology applications. The IOT methods have recently demonstrated great potential for monitoring the quality of continuous powder mixtures by near-infrared (NIR) spectroscopy. However, the dynamic conditions of continuous manufacturing processes may limit the effectiveness of such approaches. Density variations introduced to NIR spectra that are collected from dynamic powder mixtures at different process conditions is detrimental to the drug prediction accuracy and robustness of IOT methods. This work introduces a new method, called external variable augmented iterative optimization technology (EVA-IOT), which incorporates the shape of non-chemical external sources of variability into the pure component spectra matrix to improve the prediction accuracy and robustness of the base IOT algorithm. This approach derives the shape of non-chemical external variables from the latent structure of decomposition methods using NIR spectra acquired from a single mixture at known levels of the external variable. A density-augmented EVA-IOT method was developed and implemented to quantify the active pharmaceutical ingredient (API) in continuous powder mixtures flowing at varying process conditions in a simulated continuous process. The EVA-IOT method demonstrated a significantly enhanced API prediction accuracy and robustness against process variation compared to alternative IOT methods. The overall prediction performance of EVA-IOT was comparable to that of global partial least square (PLS) regression models while reducing the calibration burden up to 97%. This makes EVA-IOT a material-sparing alternative to calibration-intensive robust decomposition modeling approaches for monitoring the quality of continuous pharmaceutical powder streams.

Unraveling clay-mineral genesis and climate change on Earth and Mars using machine learning-based VNIR spectral modeling

Abstract Clay minerals are common in martian geological units and are globally widespread on Earth. Understanding the origin, formation, and alteration of clay minerals is crucial for unraveling past environmental conditions on Earth and Mars, in which the composition and crystallinity of clay minerals serve as important surrogate indicators for addressing these issues. Here, 621 soil and sediment samples from five chronosequences representing different climatic zones of China were investigated using visible to near-infrared reflectance (VNIR) in combination with X-ray diffraction (XRD) analysis. The crystallinity of clay minerals (i.e., illite crystallinity, illite chemistry index, kaolinite crystallinity) and clay mineral alteration index (CMAI) were analyzed with conventional methods and then predicted through a spectral modeling approach. Our results show that kaolinite with a pedogenic or sedimentary origin is characterized by a broad crystallinity range and a poorly ordered structure, especially when generated in an intense weathering environment. Predictive models were constructed with data-mining methods, including partial least-squares regression (PLSR), random forest (RF), and Cubist algorithms. The predictive performance of the crystallinity and CMAI proxies is robust, with an overall accuracy of 78% and a residual prediction deviation (RPD) of 2.57. We also found that the model’s accuracy in predicting clay-mineral-related proxies increased by 45% using random forest (RF) and Cubist compared to the PLSR models. We suggest that VNIR spectroscopy combined with RF and Cubist methods has the potential to be an alternative and broadly applicable tool for analyzing typical clay-mineral proxies, substituting for a series of common mineralogic analyses. Spectral modeling can reveal genetic and climatic information at both field and regional scales, which has profound implications for Mars missions and other space exploration programs.

A comparative study on rapid qualitative and quantitative determination of olive oil adulteration.

Authenticity of olive oil is a significant concern for producers, consumers, and policymakers. To help address this issue, a rapid, efficient, and accurate flow injection mass spectrometric (FIMS) fingerprinting approach, combined with SVM and PLS classification and regression models, was proposed for the identification and quantitative analysis of olive oil adulteration. Based on the comprehensive comparative analysis, SVM outperformed those of PLS-DA, achieving higher values for accuracy, sensitivity, and specificity, as well as positive predictive and negative predictive values, in identifying adulterated olive oil samples. Furthermore, compared with PLSR model, the SVR model demonstrated superior performance in determining the content of adulterated olive oil, with a higher coefficient of determination and lower Root Mean Square Error. In conclusion, FIMS fingerprinting technology in combination with SVM can be effectively implemented for rapid, reliable, and accurate identification and quantification of olive oil adulteration.

Comparison of individualized facial growth prediction models using artificial intelligence and partial least squares based on the Mathews growth collection.

To develop facial growth prediction models using artificial intelligence (AI) under various conditions, and to compare performance of these models with each other as well as with the partial least squares (PLS) growth prediction model. Longitudinal lateral cephalograms from 33 subjects in the Mathews growth collection were utilized. A total of 1257 pairs of before and after growth lateral cephalograms were included. In each image, 46 hard and 32 soft tissue landmarks were manually identified. Growth prediction models were constructed using a deep learning method based on TabNet deep neural network and partial least squares (PLS) method. Prediction accuracies of the two methods were compared. On average, artificial intelligence (AI) showed 0.61 mm less prediction error than PLS. Among the 77 predicted landmarks, AI was more accurate than PLS in 60 landmarks. When comparing AI models with varying numbers of training epochs, those with higher epochs yielded more accurate predictions. Overall, PLS and AI exhibited greater prediction errors for soft tissue and mandibular landmarks compared to hard tissue and maxillary landmarks. However, AI showed a smaller increase in prediction error in areas with greater variability. AI proved to be a valuable growth prediction method, with clinically acceptable prediction errors averaging 1.49 mm for 45 hard tissue landmarks and 1.71 mm for 32 soft tissue landmarks. PLS accurately predicted landmarks with low variability. However, AI generally outperformed PLS, particularly for landmarks in the lower part of the craniofacial structure and soft tissue, where uncertainty is considerable.

Neighbourhood Component Regression Approach for Housing Unit Price Prediction

Predicting housing unit price (HUP) is important for potential buyers and investors to make informed decisions. This study proposes a novel HUP prediction model based on neighbourhood component regression (NCR). The proposed NCR model was compared with other competitive methods such as principal component regression (PCR), multiple linear regression (MLR), partial least squares regression (PLSR), and generalised linear model (GLM). When tested with real datasets, the proposed NCR method revealed prediction superiority over the four state-of-the-art methods (PCR, MLR, PLSR, and GLM). This was evident from the Mean Absolute Percentage Error (MAPE), Correlation Coefficient (R), Scatter Index (SI), and Percentage Root Mean Square Error (PRMSE) utilised as model evaluation metrics. The results revealed that the NCR model had the lowest MAPE (0.0977), SI (0.0011), PRMSE (0.1130), and highest R (0.9999) as compared with the other investigated methods. This confirms the proposed NCR method’s strength for efficient and reliable HUP prediction.

Bridging Work and Family Life: The Role of Work Engagement in Optimizing Job Crafting at Bringkoning Health Center

Research Aims: The aim of this research is to analyze the role of work engagement in mediating the relationship between job crafting and work family enrichment. The population selected in this study were employees of UPTD PUSKESMAS Bringkoning, Banyuates District, Sampang Regency. Design/methodology/approach: The sampling technique used is purposive sampling, namely taking sample members from the population by determining certain criteria. The sample from this research were employees of UPTD PUSKESMAS Bringkoning, Banyuates District, Sampang Regency, who were 36 permanent employees. The data analysis method uses partial least squares (PLS). Research Findings: The results of this research prove that there is a direct influence of job crafting , work engagement on work family enrichment and there is an indirect influence of job crafting on work family enrichment and work engagement is able to mediate the relationship between job crafting and work family enrichment for employees of UPTD PUSKESMAS Bringkoning, Banyuates District, Sampang Regency. However, the direct relationship is higher when compared to the indirect influence Theoretical Contribution/Originality: By bridging these concepts, this research explores a deeper understanding of how changes in an individual's work can have a positive impact on their family life. The integration of the concepts of job crafting and work-family enrichment, especially in the context of health services, provides new insights into the dynamics of the relationship between work and personal life. The findings of this research have significant implications for human resource management in developing policies and practices that support employee well-being and balance between work life and family life.

Store Novelty dan Loyalitas Toko: Pengaruh Mediasi Store Design Complexity dan Store Design Pleasure

One of the biggest challenges faced by the retail sector in the digital era is maintaining customer loyalty to stores. This study aims to analyze the effect of store novelty on store loyalty, with store design complexity and store design pleasure as mediating variables. The study employed a purposive sampling method, involving 120 respondents who were customers of fashion stores in Surabaya. Data collection was conducted using a questionnaire distributed online via Google Forms. The collected data were analyzed using the Partial Least Square (PLS) technique with the SmartPLS 3.0 application. The results show that store novelty does not have a direct effect on store loyalty; however, it has an indirect effect mediated by store design complexity and store design pleasure. These findings provide significant contributions to the retail management literature, particularly regarding strategies for building and maintaining customer loyalty. From a managerial perspective, the study recommends that store managers focus on innovative designs emphasizing variety and comfort to create a unique and enjoyable shopping experience for customers. Effectively managing engaging and complex store design elements can serve as a powerful strategy to enhance customer loyalty amidst the increasing competition in the retail industry.

A Study on Factors Influencing Behavioral Intention to Adopt and Use Neobanking

Financial inclusion has a significant impact oneconomic growth and development and improves the livingstandards of individuals in emerging economies. After the impact of demonetization, the tendency among people to incline towards digital payments and net banking increased. Digital payment apps, e-wallets, and online shopping are now going through a higher growth rate. Neobanking is a concept completely powered by Fintech wherein there exists no bank in its physical form and complete operation is in its virtual form. This has enabled the Neobanking system to be the financial institution in its complete sustainable potential. The research is carried out to study the Behavioral Intention to adopt and use Neobanks with a sample size of 131 respondents randomly. The research deals with thearea of Financial Technology and Behavioural Finance. As a part of the research, a developed questionnaire was circulated. After collecting the data, Reliability tests such as Cronbach's alpha and validity tests such as Average Variance extracted were conducted to understand the reliability of the constructs used. Partial Least Squares (PLS) - Structural Equation Modelling (SEM) analysis and bootstrapping analysis are conducted to understand the impact of constructs on Behaviour Intention to adopt and use Neobanks. The results identified that among the Perceived Risks, Perceived Financial Risk (PFR) and Perceived Privacy Risk (PPR) have a strong, positive and significant impact on Behavioral Intention to Adopt (BIA) Neobanks. Among the key constructs of the UTAUT model only Hedonic Motivation (HM) has a significant impact on Behavioral Intention to use (BIU) Neobanks. Demographic factors further helped us to understand the Behavioral Intention to Adopt and Use Neobanks.

Use of Fourier transform infrared spectroscopy and chemometric analysis to predict the bacilloscopic index in the saliva of leprosy patients.

Leprosy is an infectious disease caused by the Mycobacterium leprae complex. Although the diagnosis of this disease is clinical, additional tests are often needed. One of the most common tests is intradermal smear bacilloscopy, which detects the patient's bacillary load. However, bacilloscopy is an invasive test and its results are subjective. The aim of this study was to propose a methodology based on Fourier transform infrared (FTIR) spectroscopy to predict the bacilloscopic index in the saliva of leprosy patients. A total of 88 saliva samples were included, and the bacilloscopic index of the patients was previously defined by intradermal smear bacilloscopy. The saliva samples were analyzed by attenuated total reflection (ATR)-FTIR; a supervised analysis with partial least squares regression (PLS), after orthogonal signal correction (OSC), was used to predict the bacilloscopic index. The PLS model with three latent variables and one OSC component obtained a coefficient of determination greater than 0.9999 in both the calibration and prediction sets. The model was able to predict the bacilloscopic indexes with a minimum error, demonstrating its robustness. Thus, modeling with ATR/FTIR and multivariate calibration using one OSC component showed high accuracy and reproducibility, overcoming the limitations of intradermal smear bacilloscopy and proving to be a promising method in the clinical management of patients with leprosy.

A Model for Diagnosing Mild Nutrient Stress in Facility-Grown Tomatoes Throughout the Entire Growth Cycle

The effective diagnosis of mild nutrient stress across the complete growth cycle of facility-grown tomatoes is challenging. This study proposes a deep learning framework based on CNN + LSTM, using canopy near-infrared spectroscopy from different growth stages of tomatoes as input, to diagnose mild stress of nitrogen (N), potassium (K), and calcium (Ca) throughout the entire growth cycle of facility-grown tomatoes. The study compares the diagnostic performance of Random Forest (RF), Support Vector Machine (SVM), Partial Least Squares (PLS), Convolutional Neural Networks (CNNs), and CNN + Long Short-Term Memory (LSTM) models for detecting mild nutrient stress in facility-grown tomatoes. Firstly, the preprocessing method of spectral characteristic bands combined with Savitzky-Golay (SG) + Standard Normal Variate (SNV) was determined. Subsequently, all sample data were divided into six groups: N-deficient, K-deficient, Ca-deficient, N-excess, K-excess, and Ca-excess. The aforementioned models were then used for classification prediction. The results show that RF and CNN + LSTM models demonstrated good predictive performance. Specifically, RF achieved accuracy rates of 70.14%, 90.81%, 88.59%, and 85.37% in the classification tasks of Ca-deficient, N-excess, K-excess, and Ca-excess, respectively. The CNN + LSTM model achieved accuracy rates of 93.33%, 63.33%, 99.2%, 83.33%, and 98.52% in the classification tasks of K-deficient, Ca-deficient, N-excess, K-excess, and Ca-excess, respectively. Finally, in the Leave-One-Group-Out Validation (LOGOV) for validating the model’s generalisation performance, RF performed better in the N-deficient, K-deficient, and Ca-deficient tasks, achieving diagnostic accuracy rates of 80.19%, 81.43%, and 77.02%, respectively. The CNN + LSTM model showed a diagnostic accuracy rate of 66.72% in the N-excess classification task. The study concludes that, given complete training data, the CNN + LSTM model can effectively diagnose mild nutrient stress (N, K, and Ca) in facility-grown tomatoes in most scenarios.