Support vector machine regression (SVR/LS-SVM)--an alternative to neural networks (ANN) for analytical chemistry? Comparison of nonlinear methods on near infrared (NIR) spectroscopy data

Leveraging Temporal Information to Improve Machine Learning-Based Calibration Techniques for Low-Cost Air Quality Sensors

Low-cost ambient sensors have been identified as a promising technology for monitoring air pollution at a high spatio-temporal resolution. However, the pollutant data captured by these cost-effective sensors are less accurate than their conventional counterparts and require careful calibration to improve their accuracy and reliability. In this paper, we propose to leverage temporal information, such as the duration of time a sensor has been deployed and the time of day the reading was taken, in order to improve the calibration of low-cost sensors. This information is readily available and has so far not been utilized in the reported literature for the calibration of cost-effective ambient gas pollutant sensors. We make use of three data sets collected by research groups around the world, who gathered the data from field-deployed low-cost CO and NO2 sensors co-located with accurate reference sensors. Our investigation shows that using the temporal information as a co-variate can significantly improve the accuracy of common machine learning-based calibration techniques, such as Random Forest and Long Short-Term Memory.

Convolutional Neural Networks Guided Raman Spectroscopy as a Process Analytical Technology (PAT) Tool for Monitoring and Simultaneous Prediction of Monoclonal Antibody Charge Variants

BACKGROUND

Charge related heterogeneities of monoclonal antibody (mAb) based therapeutic products are increasingly being considered as a critical quality attribute (CQA). They are typically estimated using analytical cation exchange chromatography (CEX), which is time consuming and not suitable for real time control. Raman spectroscopy coupled with artificial intelligence (AI) tools offers an opportunity for real time monitoring and control of charge variants.

OBJECTIVE

We present a process analytical technology (PAT) tool for on-line and real-time charge variant determination during process scale CEX based on Raman spectroscopy employing machine learning techniques.

METHOD

Raman spectra are collected from a reference library of samples with distribution of acidic, main, and basic species from 0-100% in a mAb concentration range of 0-20 g/L generated from process-scale CEX. The performance of different machine learning techniques for spectral processing is compared for predicting different charge variant species.

RESULT

A convolutional neural network (CNN) based model was successfully calibrated for quantification of acidic species, main species, basic species, and total protein concentration with R2 values of 0.94, 0.99, 0.96 and 0.99, respectively, and the Root Mean Squared Error (RMSE) of 0.1846, 0.1627, and 0.1029 g/L, respectively, and 0.2483 g/L for the total protein concentration.

CONCLUSION

We demonstrate that Raman spectroscopy combined with AI-ML frameworks can deliver rapid and accurate determination of product related impurities. This approach can be used for real time CEX pooling decisions in mAb production processes, thus enabling consistent charge variant profiles to be achieved.

Estimation of the distribution patterns of heavy metal in soil from airborne hyperspectral imagery based on spectral absorption characteristics

Though soil is widely known as one of the most valuable resources for the world, its quality is going to be lower because of unsustainable economic development and social progress. Therefore, it is important for us to monitor and evaluate the quality of soil, especially its heavy metal contents which is too scarce to identify in soil spectra easily but poisonous enough to affect human health in a long run. Most of the existing estimation methods have based the characteristic bands on statistical analysis to a large extent, which is hard to accurately explain the retrieval mechanism. In this paper, the absorption characteristics of heavy metal are studied based on the soil spectra, and the distribution pattern is mapped in a large-scale continuous space, for environmental monitoring and further decision support. Taking Yitong County, China as the study area. After spectra continuum removal, the heavy metal contents were estimated by 11 features including the absorption depth, absorption area, and band ratio around 2200 nm, which showed the best performance. For arsenic (As), the best model yields Rp2 value of 0.8474, and the RMSEP value is 36.1542 (mg/kg). It is concluded that As is adsorbed by organic matter, clay minerals, and iron/manganese oxides in soil, and the adsorption of As by first two components is greater than that of the last. For airborne spectra after continuum removal, combining the spectral absorption characteristic parameters and the highly correlated bands is more accurate than using the spectral absorption characteristic parameters or bands alone. AdaBoost is presented for the heavy metal estimation, and the fitting ability of the method is found to be stronger than that of the traditional classical methods, with the Rp2 values of 0.6242 and the RMSEP value of 43.6481 (mg/kg). In summary, these results will provide a prospective basis for the rapid estimation of soil heavy metals, the risk assessment of soil heavy metals and soil environmental monitoring in a large scale.

Prediction of Thermogravimetric Data in the Thermal Recycling of e-waste Using Machine Learning Techniques: A Data-driven Approach

The release of bromine-free hydrocarbons and gases is a major challenge faced in the thermal recycling of e-waste due to the corrosive effects of produced HBr. Metal oxides such as Fe2O3 (hematite) are excellent debrominating agents, and they are copyrolyzed along with tetrabromophenol (TBP), a lesser used brominated flame retardant that is a constituent of printed circuit boards in electronic equipment. The pyrolytic (N2) and oxidative (O2) decomposition of TBP with Fe2O3 has been previously investigated with thermogravimetric analysis (TGA) at four different heating rates of 5, 10, 15, and 20 °C/min, and the mass loss data between room temperature and 800 °C were reported. The objective of our paper is to study the effectiveness of machine learning (ML) techniques to reproduce these TGA data so that the use of the instrument can be eliminated to enhance the potential of online monitoring of copyrolysis in e-waste treatment. This will reduce experimental and human errors as well as improve process time significantly. TGA data are both nonlinear and multidimensional, and hence, nonlinear regression techniques such as random forest (RF) and gradient boosting regression (GBR) showed the highest prediction accuracies of 0.999 and lowest prediction errors among all the ML models employed in this work. The large data sets allowed us to explore three different scenarios of model training and validation, where the number of training samples were varied from 10,000 to 40,000 for both TBP and TBP + hematite samples under N2 (pyrolysis) and O2 (combustion) environments. The novelty of our study is that ML techniques have not been employed for the copyrolysis of these compounds, while the significance is the excellent potential of enhanced online monitoring of e-waste treatment and extension to other characterization techniques such as spectroscopy and chromatography. Lastly, e-waste recycling could greatly benefit from ML applications since it has the potential to reduce total and operational costs and improve overall process time and efficiency, thereby encouraging more treatment plants to adopt these techniques, resulting in reducing the increasing environmental footprint of e-waste.

Recent developments of e-sensing devices coupled to data processing techniques in food quality evaluation: a critical review

A greater demand for high-quality food is being driven by the growth of economic and technological advancements. In this context, consumers are currently paying special attention to organoleptic characteristics such as smell, taste, and appearance. Motivated to mimic human senses, scientists developed electronic devices such as e-noses, e-tongues, and e-eyes, to spot signals relative to different chemical substances prevalent in food systems. To interpret the information provided by the sensors' responses, multiple chemometric approaches are used depending on the aim of the study. This review based on the Web of Science database, endeavored to scrutinize three e-sensing systems coupled to chemometric approaches for food quality evaluation. A total of 122 eligible articles pertaining to the e-nose, e-tongue and e-eye devices were selected to conduct this review. Most of the performed studies used exploratory analysis based on linear factorial methods, while classification and regression techniques came in the second position. Although their applications have been less common in food science, it is to be noted that nonlinear approaches based on artificial intelligence and machine learning deployed in a big-data context have generally yielded better results for classification and regression purposes, providing new perspectives for future studies.

Machine-learning-facilitated prediction of heavy metal contamination in distiller's dried grains with solubles

Excessive heavy metal contamination often occurs in feed due to natural or anthropogenic activity, leading to poisoning and other health problems in animals. In this study, a visible/near-infrared hyperspectral imaging system (Vis/NIR HIS) was used to reveal the different characteristics of spectral reflectance of Distillers Dried Grains with Solubles (DDGS) doped with various heavy metals and to effectively predict metal concentrations. Two types of sample treatment were used, namely tablet and bulk. Three quantitative analysis models were constructed based on the full wavelength, and through comparison the support vector regression (SVR) model was found to show the best performance. As typical heavy metal contaminants, copper (Cu) and zinc (Zn) were used for modeling and prediction. The prediction set accuracy of the tablet samples doped with Cu and Zn was 94.9% and 86.2%, respectively. In addition, a novel characteristic wavelength selection model based on SVR (SVR-CWS) was proposed to filter characteristic wavelengths, which improved the detection performance. The regression accuracy of the SVR model on the prediction set of tableted samples with different Cu and Zn concentrations was 94.7% and 85.9%, respectively. The accuracy of bulk samples with different Cu and Zn concentrations was 81.3% and 80.3%, respectively, which indicated that the detection method can reduce the pretreatment steps and verify its practicability. The overall results suggested the potential of Vis/NIR-HIS in the detection of feed safety and quality.

High-Dimensional Data Global Sensitivity Analysis Based on Deep Soft Sensor Model

This article investigates the sensitivity analysis (SA) of high-dimensional data to identify the effects of process variables on output quantity of interest (QoI) in industrial soft sensor modeling. The computational cost of analyzing the SA of high-dimensional data is high, and models available for SA techniques usually have limited generalization capacity. Therefore, we propose a novel high-dimensional data global SA (GSA) approach based on a deep soft sensor model to address these issues. We first develop an approximately incremental grouping (AIG) algorithm and a region-based cooperative co-evolution (RBCC) algorithm to decompose the high-dimensional data into independent regions for the GSA. Subsequently, a multihead deep soft sensor model with generalization performance is designed to determine the GSA indices of each decomposed region. Specifically, the region of interest (RoI) align algorithm provides the multihead with precisely located decomposed region features. Finally, based on the uncertainty analysis of each model head, we present a joint loss function with the Monte Carlo dropout (MC-dropout) algorithm to measure the GSA indices of each decomposed region on QoIs. Experimental evaluation results on a benchmark dataset and a real-world one demonstrate the effectiveness of the proposed approach in addressing the GSA of high-dimensional data in industrial processes.

A Novel Approach for Therapeutic Drug Monitoring of Valproic Acid Using FT-IR Spectroscopy and Nonlinear Support Vector Regression

BACKGROUND

Recent technological progress has bolstered efforts to bring personalized medicine from theory into clinical practice. However, progress in areas such as therapeutic drug monitoring (TDM) has remained somewhat stagnant. In drugs with well-known dose-response relationships, TDM can enhance patient outcomes and reduce health care costs. Traditional monitoring methods such as chromatography-based or immunoassay techniques are limited by their higher costs and slow turnaround times, making them unsuitable for real-time or onsite analysis.

OBJECTIVE

In this work, we propose the use of a fast, direct, and simple approach using Fourier transform infrared spectroscopy (FT-IR) combined with chemometric techniques for the therapeutic monitoring of valproic acid (VPA).

METHOD

In this context, a database of FT-IR spectra was constructed from human plasma samples containing various concentrations of VPA; these samples were characterized by the reference method (immunoassay technique) to determine the VPA contents. The FT-IR spectra were processed by two chemometric regression methods: partial least-squares regression (PLS) and support vector regression (SVR).

RESULTS

The results provide good evidence for the effectiveness of the combination of FT-IR spectroscopy and SVR modeling for estimating VPA in human plasma. SVR models showed better predictive abilities than PLS models in terms of root-mean-square error of calibration and prediction RMSEC, RMSEP, R2Cal, R2Pred, and residual predictive deviation (RPD).

CONCLUSIONS

This analytical tool offers potential for real-time TDM in the clinical setting.

HIGHLIGHTS

FTIR spectroscopy was evaluated for the first time to predict VPA in human plasma for TDM. Two regressions were evaluated to predict VPA in human plasma, and the best-performing model was obtained using nonlinear SVR.

An applicability test of the conventional and neural network methods to map the overall water quality of the Caspian Sea

This study investigates the water quality of the Caspian Sea by examining the presence of nutrients and heavy metals in the water. Water samples were collected from 22 stations and analyzed for nutrient and heavy metal levels. The study used the fuzzy method to prepare water quality maps and employed ANNs methods to predict microbial contamination for future years. The results revealed that the western and northwestern parts of the region had higher nutrient levels (about 40.2 % of the region), while the eastern and northeastern shores were highly polluted due to increased urbanization (about 70.1 % of the region). The long short-term memory (LSTM) method was found to have the highest accuracy compared to other ANNs methods and indicated a recent increase in pollution (R_{Water quality}²=0.940, R_OECD²=0.950, R_TRIX²=0.840). The study recommends targeted research to identify the causes and means of controlling pollution in light of the predicted increase in pollution in the Caspian Sea.

Identification of gallbladder cancer by direct near-infrared measurement of raw bile combined with two-trace two-dimensional correlation analysis

We demonstrated the utility of direct near-infrared (NIR) bile analysis for the identification of gallbladder (GB) cancer by employing two-trace two-dimensional (2T2D) correlation analysis to recognize dissimilar spectral features among diverse bile samples for potential improvement of discrimination accuracy. To represent more diverse clinical cases for reliable assessment, bile samples obtained from five normal, 44 gallstone, 25 GB polyp, six hepatocellular cancer (HCC), and eight GB cancer subjects were analyzed. Due to the altered metabolic pathways by carcinogenesis, the NIR spectral features of GB cancer samples, including intensity ratios of main peaks, were different from those of other sample groups. The differentiation of GB cancer in the principal component (PC) score domain was mediocre and subsequent discrimination accuracy based on linear discriminant analysis (LDA) was 88.5%. When 2T2D slice spectra were obtained using a reference spectrum constructed by the linear combination of the spectra of five pure representative bile metabolites and employed, the accuracy was improved to 95.6%. The sensitive recognition of dissimilar spectral features in GB cancer by 2T2D correlation analysis was responsible for the enhanced discrimination.

Machine Learning-Based Analytical Systems: Food Forensics

Despite a large amount of money being spent on both food analyses and control measures, various food-borne illnesses associated with pathogens, toxins, pesticides, adulterants, colorants, and other contaminants pose a serious threat to human health, and thus food safety draws considerable attention in the modern pace of the world. The presence of various biogenic amines in processed food have been frequently considered as the primary quality parameter in order to check food freshness and spoilage of protein-rich food. Various conventional detection methods for detecting hazardous analytes including microscopy, nucleic acid, and immunoassay-based techniques have been employed; however, recently, array-based sensing strategies are becoming popular for the development of a highly accurate and precise analytical method. Array-based sensing is majorly facilitated by the advancements in multivariate analytical techniques as well as machine learning-based approaches. These techniques allow one to solve the typical problem associated with the interpretation of the complex response patterns generated in array-based strategies. Consequently, the machine learning-based neural networks enable the fast, robust, and accurate detection of analytes using sensor arrays. Thus, for commercial applications, most of the focus has shifted toward the development of analytical methods based on electrical and chemical sensor arrays. Therefore, herein, we briefly highlight and review the recently reported array-based sensor systems supported by machine learning and multivariate analytics to monitor food safety and quality in the field of food forensics.

Quantitative study of impact damage on yellow peaches based on reflectance, absorbance and Kubelka-Munk spectral data

Impact damage is one of the main forms of damage during the postharvest transportation and processing of yellow peaches. Thus, a quantitative prediction of the impact damage degree of yellow peaches is significant for their postharvest grading. In the present study, mechanical parameters such as the damage area, absorbed energy and maximum force were obtained based on a single pendulum collision device and an intelligent data acquisition system. The reflection spectra (R) of the damaged areas of yellow peaches were collected by a hyperspectral imaging system and transformed into absorbance (A) spectra and Kubelka-Munk (K-M) spectra. The R, A and K-M spectra were preprocessed by standard normal variables (SNV), moving average (MA) and Gaussian filtering (GF). Partial least squares regression (PLSR) models and support vector regression (SVR) models based on original and preprocessed spectra were established, respectively. By comparative analysis, the spectral data with better prediction performance (raw or preprocessed spectra) were selected from all spectra, and the characteristic wavelengths were selected by competitive adaptive reweighted sampling (CARS) and uninformative variable elimination (UVE). The PLSR and SVR models based on characteristic wavelengths were established. The results revealed that the prediction performance of the K-M-GF-CARS-PLSR model is the best. For the damage area, absorbed energy and maximum force, the R _P ² and RMSEP of the K-M-GF-CARS-PLSR model were 0.870 and 77.865 mm², 0.772 and 1.065 J, 0.895 and 47.996 N, respectively. Furthermore, the values of their RPD were 2.700, 1.768 and 3.050, respectively. The characteristic wavelengths of the model were 18.8%, 10.2% and 21.6%, respectively. The results of this study showed that there was a strong correlation between the mechanical parameters and K-M spectrum, which demonstrates the feasibility of quantitatively predicting the damage degree of yellow peaches based on the K-M spectrum. Therefore, the results of this work not only provide theoretical guidance for the postharvest grading of fruits, but also enrich the theoretical system of biomechanics.

Applications of Artificial Intelligence and Machine Learning Algorithms to Crystallization

Artificial intelligence and specifically machine learning applications are nowadays used in a variety of scientific applications and cutting-edge technologies, where they have a transformative impact. Such an assembly of statistical and linear algebra methods making use of large data sets is becoming more and more integrated into chemistry and crystallization research workflows. This review aims to present, for the first time, a holistic overview of machine learning and cheminformatics applications as a novel, powerful means to accelerate the discovery of new crystal structures, predict key properties of organic crystalline materials, simulate, understand, and control the dynamics of complex crystallization process systems, as well as contribute to high throughput automation of chemical process development involving crystalline materials. We critically review the advances in these new, rapidly emerging research areas, raising awareness in issues such as the bridging of machine learning models with first-principles mechanistic models, data set size, structure, and quality, as well as the selection of appropriate descriptors. At the same time, we propose future research at the interface of applied mathematics, chemistry, and crystallography. Overall, this review aims to increase the adoption of such methods and tools by chemists and scientists across industry and academia.

A review of the application of machine learning in water quality evaluation

With the rapid increase in the volume of data on the aquatic environment, machine learning has become an important tool for data analysis, classification, and prediction. Unlike traditional models used in water-related research, data-driven models based on machine learning can efficiently solve more complex nonlinear problems. In water environment research, models and conclusions derived from machine learning have been applied to the construction, monitoring, simulation, evaluation, and optimization of various water treatment and management systems. Additionally, machine learning can provide solutions for water pollution control, water quality improvement, and watershed ecosystem security management. In this review, we describe the cases in which machine learning algorithms have been applied to evaluate the water quality in different water environments, such as surface water, groundwater, drinking water, sewage, and seawater. Furthermore, we propose possible future applications of machine learning approaches to water environments.

Comprehensive Modeling in Predicting Liquid Density of the Refrigerant Systems Using Least-Squares Support Vector Machine Approach

A robust machine learning algorithm known as the least-squares support vector machine (LSSVM) model was used to predict the liquid densities of 48 different refrigerant systems. Hence, a massive dataset was gathered using the reports published previously. The proposed model was evaluated via various analyses. Based on the statistical analysis results, the actual values predicted by this model have high accuracy, and the calculated values of RMSE, MRE, STD, and R2 were 0.0116, 0.158, 0.1070, and 0.999, respectively. Moreover, sensitivity analysis was done on the efficient input parameters, and it was found that CF2H2 has the most positive effect on the output parameter (with a relevancy factor of +50.19). Furthermore, for checking the real data accuracy, the technique of leverage was considered, the results of which revealed that most of the considered data are reliable. The power and accuracy of this simple model in predicting liquid densities of different refrigerant systems are high; therefore, it is an appropriate alternative for laboratory data.

Multi-Temporal Landsat-8 Images for Retrieval and Broad Scale Mapping of Soil Copper Concentration Using Empirical Models

Mapping soil heavy metal concentration using machine learning models based on readily available satellite remote sensing images is highly desirable. Accurate mapping relies on appropriate data, feature extraction, and model selection. To this end, a data processing pipeline for soil copper (Cu) concentration estimation has been designed. First, instead of using single Landsat scenes, the utilization of multiple Landsat scenes of the same location over time is considered. Second, to generate a preferred feature set as input to a regression model, a number of feature extraction methods are motivated and compared. Third, to find a preferred regression model, a variety of approaches are implemented and compared for accuracy. In this research, 11 Landsat-8 images from 2013 to 2017 of Gulin County, Sichuan China, and 138 soil samples with lab-measured Cu concentrations collected from the area in 2015 are used. A variety a metrics under cross-validation are used for comparison. The results indicate that multi-temporal images increase accuracy compared to single Landsat images. The preferred feature extraction varies based on the regression model used; however, the best results are obtained using support vector regression and the original data. The final soil Cu map generated using the recommended data processing pipeline shows a consistent spatial pattern with a ground-truth land cover classification map. These results indicate that machine learning has the ability to perform large-scale soil heavy metal mapping from widely available satellite remote sensing images.

A new alternative tool to analyse glycosylation in pharmaceutical proteins based on infrared spectroscopy combined with nonlinear support vector regression

Almost 60% of commercialized pharmaceutical proteins are glycosylated. Glycosylation is considered a critical quality attribute, as it affects the stability, bioactivity and safety of proteins. Hence, the development of analytical methods to characterise the composition and structure of glycoproteins is crucial. Currently, existing methods are time-consuming, expensive, and require significant sample preparation steps, which can alter the robustness of the analyses. In this work, we suggest the use of a fast, direct, and simple Fourier transform infrared spectroscopy (FT-IR) combined with a chemometric strategy to address this challenge. In this context, a database of FT-IR spectra of glycoproteins was built, and the glycoproteins were characterised by reference methods (MALDI-TOF, LC-ESI-QTOF and LC-FLR-MS) to estimate the mass ratio between carbohydrates and proteins and determine the composition in monosaccharides. The FT-IR spectra were processed first by Partial Least Squares Regression (PLSR), one of the most used regression algorithms in spectroscopy and secondly by Support Vector Regression (SVR). SVR has emerged in recent years and is now considered a powerful alternative to PLSR, thanks to its ability to flexibly model nonlinear relationships. The results provide clear evidence of the efficiency of the combination of FT-IR spectroscopy, and SVR modelling to characterise glycosylation in therapeutic proteins. The SVR models showed better predictive performances than the PLSR models in terms of RMSECV, RMSEP, R2CV, R2Pred and RPD. This tool offers several potential applications, such as comparing the glycosylation of a biosimilar and the original molecule, monitoring batch-to-batch homogeneity, and in-process control.

Towards the direct detection of viral materials at the surface of protective face masks via infrared spectroscopy

The ongoing COVID-19 pandemic represents a considerable risk for the general public and especially for health care workers. To avoid an overloading of the health care system and to control transmission chains, the development of rapid and cost-effective techniques allowing for the reliable diagnosis of individuals with acute respiratory infections are crucial. Uniquely, the present study focuses on the development of a direct face mask sampling approach, as worn (i.e., used) disposable face masks contain exogenous environmental constituents, as well as endogenously exhaled breath aerosols. Optical techniques—and specifically infrared (IR) molecular spectroscopic techniques—are promising tools for direct virus detection at the surface of such masks. In the present study, a rapid and non-destructive approach for monitoring exposure scenarios via medical face masks using attenuated total reflection infrared spectroscopy is presented. Complementarily, IR external reflection spectroscopy was evaluated in comparison for rapid mask analysis. The utility of a face mask-based sampling approach was demonstrated by differentiating water, proteins, and virus-like particles sampled onto the mask. Data analysis using multivariate statistical algorithms enabled unambiguously classifying spectral signatures of individual components and biospecies. This approach has the potential to be extended towards the rapid detection of SARS-CoV-2—as shown herein for the example of virus-like particles which are morphologically equivalent to authentic virus—without any additional sample preparation or elaborate testing equipment at laboratory facilities. Therefore, this strategy may be implemented as a routine large-scale monitoring routine, e.g., at health care institutions, nursing homes, etc. ensuring the health and safety of medical personnel.

Machine Learning and Statistical Prediction of Pitching Arm Kinetics

BACKGROUND

Over the past decade, research has attempted to elucidate the cause of throwing-related injuries in the baseball athlete. However, when considering the entire kinetic chain, full body mechanics, and pitching cycle sequencing, there are hundreds of variables that could influence throwing arm health, and there is a lack of quality investigations evaluating the relationship and influence of multiple variables on arm stress.

PURPOSE

To identify which variables have the most influence on elbow valgus torque and shoulder distraction force using a statistical model and a machine learning approach.

STUDY DESIGN

Cross-sectional study; Level of evidence, 3.

METHODS

A retrospective review was performed on baseball pitchers who underwent biomechanical evaluation at the university biomechanics laboratory. Regression models and 4 machine learning models were created for both elbow valgus torque and shoulder distraction force. All models utilized the same predictor variables, which included pitch velocity and 17 pitching mechanics.

RESULTS

The analysis included a total of 168 high school and collegiate pitchers with a mean age of 16.7 years (SD, 3.2 years) and BMI of 24.4 (SD, 1.2). For both elbow valgus torque and shoulder distraction force, the gradient boosting machine models demonstrated the smallest root mean square errors and the most precise calibrations compared with all other models. The gradient boosting model for elbow valgus torque reported the highest influence for pitch velocity (relative influence, 28.4), with 5 mechanical variables also having significant influence. The gradient boosting model for shoulder distraction force reported the highest influence for pitch velocity (relative influence, 20.4), with 6 mechanical variables also having significant influence.

CONCLUSION

The gradient boosting machine learning model demonstrated the best overall predictive performance for both elbow valgus torque and shoulder distraction force. Pitch velocity was the most influential variable in both models. However, both models also revealed that pitching mechanics, including maximum humeral rotation velocity, shoulder abduction at foot strike, and maximum shoulder external rotation, significantly influenced both elbow and shoulder stress.

CLINICAL RELEVANCE

The results of this study can be used to inform players, coaches, and clinicians on specific mechanical variables that may be optimized to mitigate elbow or shoulder stress that could lead to throwing-related injury.

Surface Pretreatments of AA5083 Aluminum Alloy with Enhanced Corrosion Protection for Cerium-Based Conversion Coatings Application: Combined Experimental and Computational Analysis

The effects of surface pretreatments on the cerium-based conversion coating applied on an AA5083 aluminum alloy were investigated using a combination of scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), polarization testing, and electrochemical impedance spectroscopy. Two steps of pretreatments containing acidic or alkaline solutions were applied to the surface to study the effects of surface pretreatments. Among the pretreated samples, the sample prepared by the pretreatment of the alkaline solution then acid washing presented higher corrosion protection (~3 orders of magnitude higher than the sample without pretreatment). This pretreatment provided a more active surface for the deposition of the cerium layer and provided a more suitable substrate for film formation, and made a more uniform film. The surface morphology of samples confirmed that the best surface coverage was presented by alkaline solution then acid washing pretreatment. The presence of cerium in the (EDS) analysis demonstrated that pretreatment with the alkaline solution then acid washing resulted in a higher deposition of the cerium layer on the aluminum surface. After selecting the best surface pretreatment, various deposition times of cerium baths were investigated. The best deposition time was achieved at 10 min, and after this critical time, a cracked film formed on the surface that could not be protective. The corrosion resistance of cerium-based conversion coatings obtained by electrochemical tests were used for training three computational techniques (artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), and support vector machine regression (SVMR)) based on Pretreatment-1 (acidic or alkaline cleaning: pH (1)), Pretreatment-2 (acidic or alkaline cleaning: pH (2)), and deposition time in the cerium bath as an input. Various statistical criteria showed that the ANFIS model (R² = 0.99, MSE = 48.83, and MAE = 3.49) could forecast the corrosion behavior of a cerium-based conversion coating more accurately than other models. Finally, due to the robust performance of ANFIS in modeling, the effect of each parameter was studied.

Deep learning-based exchange rate prediction during the COVID-19 pandemic

This study proposes an ensemble deep learning approach that integrates Bagging Ridge (BR) regression with Bi-directional Long Short-Term Memory (Bi-LSTM) neural networks used as base regressors to become a Bi-LSTM BR approach. Bi-LSTM BR was used to predict the exchange rates of 21 currencies against the USD during the pre-COVID-19 and COVID-19 periods. To demonstrate the effectiveness of our proposed model, we compared the prediction performance with several more traditional machine learning algorithms, such as the regression tree, support vector regression, and random forest regression, and deep learning-based algorithms such as LSTM and Bi-LSTM. Our proposed ensemble deep learning approach outperformed the compared models in forecasting exchange rates in terms of prediction error. However, the performance of the model significantly varied during non-COVID-19 and COVID-19 periods across currencies, indicating the essential role of prediction models in periods of highly volatile foreign currency markets. By providing an improved prediction performance and identifying the most seriously affected currencies, this study is beneficial for foreign exchange traders and other stakeholders in that it offers opportunities for potential trading profitability and for reducing the impact of increased currency risk during the pandemic.

Comparison of partial least square, artificial neural network, and support vector regressions for real-time monitoring of CHO cell culture processes using in situ near-infrared spectroscopy

The biopharmaceutical industry must guarantee the efficiency and biosafety of biological medicines, which are quite sensitive to cell culture process variability. Real-time monitoring procedures based on vibrational spectroscopy such as near-infrared (NIR) spectroscopy, are then emerging to support innovative strategies for retro-control of key parameters as substrates and by-product concentration. Whereas monitoring models are mainly constructed using partial least squares regression (PLSR), spectroscopic models based on artificial neural networks (ANNR) and support vector regression (SVR) are emerging with promising results. Unfortunately, analysis of their performance in cell culture monitoring has been limited. This study was then focused to assess their performance and suitability for the cell culture process challenges. PLSR had inferior values of the determination coefficient (R² ) for all the monitored parameters (i.e., 0.85, 0.93, and 0.98, respectively for the PLSR, SVR, and ANNR models for glucose). In general, PLSR had a limited performance while models based on ANNR and SVR have been shown superior due to better management of inter-batch heterogeneity and enhanced specificity. Overall, the use of SVR and ANNR for the generation of calibration models enhanced the potential of NIR spectroscopy as a monitoring tool.

Predicting moisture content during maize nixtamalization using machine learning with NIR spectroscopy

Moisture content during nixtamalization can be accurately predicted from NIR spectroscopy when coupled with a support vector machine (SVM) model, is strongly modulated by the environment, and has a complex genetic architecture. Lack of high-throughput phenotyping systems for determining moisture content during the maize nixtamalization cooking process has led to difficulty in breeding for this trait. This study provides a high-throughput, quantitative measure of kernel moisture content during nixtamalization based on NIR scanning of uncooked maize kernels. Machine learning was utilized to develop models based on the combination of NIR spectra and moisture content determined from a scaled-down benchtop cook method. A linear support vector machine (SVM) model with a Spearman's rank correlation coefficient of 0.852 between wet laboratory and predicted values was developed from 100 diverse temperate genotypes grown in replicate across two environments. This model was applied to NIR spectra data from 501 diverse temperate genotypes grown in replicate in five environments. Analysis of variance revealed environment explained the highest percent of the variation (51.5%), followed by genotype (15.6%) and genotype-by-environment interaction (11.2%). A genome-wide association study identified 26 significant loci across five environments that explained between 5.04% and 16.01% (average = 10.41%). However, genome-wide markers explained 10.54% to 45.99% (average = 31.68%) of the variation, indicating the genetic architecture of this trait is likely complex and controlled by many loci of small effect. This study provides a high-throughput method to evaluate moisture content during nixtamalization that is feasible at the scale of a breeding program and provides important information about the factors contributing to variation of this trait for breeders and food companies to make future strategies to improve this important processing trait.

Foot Position Measurement during Assistive Motion for Sit-to-Stand Using a Single Inertial Sensor and Shoe-Type Force Sensors

Assistive motion for sit-to-stand causes lower back pain (LBP) among caregivers. Considering previous studies that showed that foot position adjustment could reduce lumbar load during assistive motion for sit-to-stand, quantitative monitoring of and instructions on foot position could contribute toward reducing LBP among caregivers. The present study proposes and evaluates a new method for the quantitative measurement of foot position during assistive motion for sit-to-stand using a few wearable sensors that are not limited to the measurement area. The proposed method measures quantitative foot position (anteroposterior and mediolateral distance between both feet) through a machine learning technique using features obtained from only a single inertial sensor on the trunk and shoe-type force sensors. During the experiment, the accuracy of the proposed method was investigated by comparing the obtained values with those from an optical motion capture system. The results showed that the proposed method produced only minor errors (less than 6.5% of body height) when measuring foot position during assistive motion for sit-to-stand. Furthermore, Bland–Altman plots suggested no fixed errors between the proposed method and the optical motion capture system. These results suggest that the proposed method could be utilized for measuring foot position during assistive motion for sit-to-stand.

From consumers' science to food functionality-Challenges and opportunities for vibrational spectroscopy

Current available methods used to measure or estimate the composition, functionality, and sensory properties of foods and food ingredients are destructive and time consuming. Therefore, new approaches are required by both the food industry and R&D organizations. Recent years have witnessed a steady growth on the applications and utilization of vibrational spectroscopy techniques [near (NIR), mid infrared (MIR), Raman] to analyse or estimate several properties in a wide range of foods and food ingredients. This chapter will provide with an overview of vibrational spectroscopy techniques, the combination of these techniques with multivariate data analysis, and examples on the use of these techniques to measure composition, and functional properties in a wide range of foods.

An empirical investigation of deviations from the Beer-Lambert law in optical estimation of lactate

The linear relationship between optical absorbance and the concentration of analytes-as postulated by the Beer-Lambert law-is one of the fundamental assumptions that much of the optical spectroscopy literature is explicitly or implicitly based upon. The common use of linear regression models such as principal component regression and partial least squares exemplifies how the linearity assumption is upheld in practical applications. However, the literature also establishes that deviations from the Beer-Lambert law can be expected when (a) the light source is far from monochromatic, (b) the concentrations of analytes are very high and (c) the medium is highly scattering. The lack of a quantitative understanding of when such nonlinearities can become predominant, along with the mainstream use of nonlinear machine learning models in different fields, have given rise to the use of methods such as random forests, support vector regression, and neural networks in spectroscopic applications. This raises the question that, given the small number of samples and the high number of variables in many spectroscopic datasets, are nonlinear effects significant enough to justify the additional model complexity? In the present study, we empirically investigate this question in relation to lactate, an important biomarker. Particularly, to analyze the effects of scattering matrices, three datasets were generated by varying the concentration of lactate in phosphate buffer solution, human serum, and sheep blood. Additionally, the fourth dataset pertained to invivo, transcutaneous spectra obtained from healthy volunteers in an exercise study. Linear and nonlinear models were fitted to each dataset and measures of model performance were compared to attest the assumption of linearity. To isolate the effects of high concentrations, the phosphate buffer solution dataset was augmented with six samples with very high concentrations of lactate between (100-600 mmol/L). Subsequently, three partly overlapping datasets were extracted with lactate concentrations varying between 0-11, 0-20 and 0-600 mmol/L. Similarly, the performance of linear and nonlinear models were compared in each dataset. This analysis did not provide any evidence of substantial nonlinearities due high concentrations. However, the results suggest that nonlinearities may be present in scattering media, justifying the use of complex, nonlinear models.

Near-Infrared Reflectance Spectroscopy for Predicting the Phospholipid Fraction and the Total Fatty Acid Composition of Freeze-Dried Beef

Research on fatty acids (FA) is important because their intake is related to human health. NIRS can be a useful tool to estimate the FA of beef but due to the high moisture and the high absorbance of water makes it difficult to calibrate the analyses. This work evaluated near-infrared reflectance spectroscopy as a tool to assess the total fatty acid composition and the phospholipid fraction of fatty acids of beef using freeze-dried meat. An average of 22 unrelated pure breed young bulls from 15 European breeds were reared on a common concentrate-based diet. A total of 332 longissimus thoracis steaks were analysed for fatty acid composition and a freeze-dried sample was subjected to near-infrared spectral analysis. 220 samples (67%) were used as a calibration set with the remaining 110 (33%) being used for validation of the models obtained. There was a large variation in the total FA concentration across the animals giving a good data set for the analysis and whilst the coefficient of variation was nearly 68% for the monounsaturated FA it was only 27% for the polyunsaturated fatty acids (PUFA). PLS method was used to develop the prediction models. The models for the phospholipid fraction had a low R²p and high standard error, while models for neutral lipid had the best performance, in general. It was not possible to obtain a good prediction of many individual PUFA concentrations being present at low concentrations and less variable than other FA. The best models were developed for Total FA, saturated FA, 9c18:1 and 16:1 with R²_p greater than 0.76. This study indicates that NIRS is a feasible and useful tool for screening purposes and it has the potential to predict most of the FA of freeze-dried beef.

Low-Cost Air Quality Sensing towards Smart Homes

The evolution of low-cost sensors (LCSs) has made the spatio-temporal mapping of indoor air quality (IAQ) possible in real-time but the availability of a diverse set of LCSs make their selection challenging. Converting individual sensors into a sensing network requires the knowledge of diverse research disciplines, which we aim to bring together by making IAQ an advanced feature of smart homes. The aim of this review is to discuss the advanced home automation technologies for the monitoring and control of IAQ through networked air pollution LCSs. The key steps that can allow transforming conventional homes into smart homes are sensor selection, deployment strategies, data processing, and development of predictive models. A detailed synthesis of air pollution LCSs allowed us to summarise their advantages and drawbacks for spatio-temporal mapping of IAQ. We concluded that the performance evaluation of LCSs under controlled laboratory conditions prior to deployment is recommended for quality assurance/control (QA/QC), however, routine calibration or implementing statistical techniques during operational times, especially during long-term monitoring, is required for a network of sensors. The deployment height of sensors could vary purposefully as per location and exposure height of the occupants inside home environments for a spatio-temporal mapping. Appropriate data processing tools are needed to handle a huge amount of multivariate data to automate pre-/post-processing tasks, leading to more scalable, reliable and adaptable solutions. The review also showed the potential of using machine learning technique for predicting spatio-temporal IAQ in LCS networked-systems.

E-sensing and nanoscale-sensing devices associated with data processing algorithms applied to food quality control: a systematic review

Devices of human-based senses such as e-noses, e-tongues and e-eyes can be used to analyze different compounds in several food matrices. These sensors allow the detection of one or more compounds present in complex food samples, and the responses obtained can be used for several goals when different chemometric tools are applied. In this systematic review, we used Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines, to address issues such as e-sensing with chemometric methods for food quality control (FQC). A total of 109 eligible articles were selected from PubMed, Scopus and Web of Science. Thus, we predicted that the association between e-sensing and chemometric tools is essential for FQC. Most studies have applied preliminary approaches like exploratory analysis, while the classification/regression methods have been less investigated. It is worth mentioning that non-linear methods based on artificial intelligence/machine learning, in most cases, had classification/regression performances superior to non-liner, although their applications were seen less often. Another approach that has generated promising results is the data fusion between e-sensing devices or in conjunction with other analytical techniques. Furthermore, some future trends in the application of miniaturized devices and nanoscale sensors are also discussed.

Using near-infrared spectroscopy to determine intramuscular fat and fatty acids of beef applying different prediction approaches

This study aimed to predict fat and fatty acids (FA) contents in beef using near-infrared spectroscopy and prediction models based on partial least squares (PLS) and support vector machine regression in radial kernel (R-SVR). Fat and FA were assessed in 200 longissimus thoracis samples, and spectra were collected in reflectance mode from ground meat. The analyses were performed for PLS and R-SVR with and without wavelength selection based on genetic algorithms (GAs). The GA application improved the error prediction by 15% and 68% for PLS and R-SVR, respectively. Models based on GA plus R-SMV showed a prediction ability for fat and FA with an average coefficient of determination of 0.92 and ratio performance deviation of 4.8.

Heavy metal Hg stress detection in tobacco plant using hyperspectral sensing and data-driven machine learning methods

Accurate detection of heavy metal stress on the growth status of plants is of great concern for agricultural production and management, food security, and ecological environment. A proximal hyperspectral imaging (HSI) system covered the visible/near-infrared (Vis/NIR) region of 400-1000 nm coupled with machine learning methods were employed to discriminate the tobacco plants stressed by different concentration of heavy metal Hg. After acquiring hyperspectral images of tobacco plants stressed by heavy metal Hg with concentration solutions of 0 mg·L-1 (non-stressed groups), 1, 3, and 5 mg·L-1 (3 stressed groups), regions of interest (ROIs) of canopy in tobacco plants were identified for spectra processing. Meanwhile, tobacco plant's appearance and microstructure of mesophyll tissue in tobacco leaves were analyzed. After that, clustering effects of the non-stressed and stressed groups were revealed by score plots and score images calculated by principal component analysis (PCA). Then, loadings of PCA and competitive adaptive reweighted sampling (CARS) algorithm were employed to pick effective wavelengths (EWs) for discriminating non-stressed and stressed samples. Partial least squares discriminant analysis (PLS-DA) and least-squares support vector machine (LS-SVM) were utilized to estimate the stressed tobacco plants status with different concentrations Hg solutions. The performances of those models were evaluated using confusion matrixes (CMes) and receiver operating characteristics (ROC) curves. Results demonstrated that PLS-DA models failed to offer relatively good result, and this algorithm was abandoned to classify the stressed and non-stressed groups of tobacco plants. Compared to LS-SVM model based on full spectra (FS-LS-SVM), the LS-SVM model established EWs selected by CARS (CARS-LS-SVM) carried 13 variables provided an accuracy of 100%, which was promising to achieve the qualitative discrimination of the non-stressed and stressed tobacco plants. Meanwhile, for revealing the discrepancy between 3 stressed groups of tobacco plants, the other FS-LS-SVM, PCA-LS-SVM, and CARS-LS-SVM models were setup and offered relatively low accuracies of 55.56%, 51.11% and 66.67%, respectively. Performance of those 3 LS-SVM discriminative models was also poorly performing to differentiate 3 stressed groups of tobacco plants, which might be caused by low concentration of heavy metal and similar canopy (especially in fresh leaves) of plant. The achievements of the research indicated that HSI coupled with machine learning methods had a powerful potential to discriminate tobacco plant stressed by heavy metal Hg.

Estimating the distribution trend of soil heavy metals in mining area from HyMap airborne hyperspectral imagery based on ensemble learning

The problem of heavy metal pollution of soils in China is severe. The traditional spectral methods for soil heavy metal monitoring and assessment cannot meet the needs for large-scale areas. Therefore, in this study, we used HyMap-C airborne hyperspectral imagery to explore the estimation of soil heavy metal concentration. Ninety five soil samples were collected synchronously with airborne image acquisition in the mining area of Yitong County, China. The pre-processed spectrum of airborne images at the sampling point was then selected by the competitive adaptive reweighted sampling (CARS) method. The selected spectral features and the heavy metal data of soil samples were inverted to establish the inversion model. An ensemble learning method based on a stacking strategy is proposed for the inversion modeling of soil samples and image data. The experimental results show that this CARS-Stacking method can better predict the four heavy metals in the study area than other methods. For arsenic (As), chromium (Cr), lead (Pb), and zinc (Zn), the determination coefficients of the test data set (R_P²) are 0.73, 0.63, 0.60, and 0.71, respectively. It was found that the estimated results and the distribution trend of heavy metals are almost the same as in actual ground measurements.

The efficacy of support vector machines in modelling deviations from the Beer-Lambert law for optical measurement of lactate

Lactate is an important biomarker with a significant diagnostic and prognostic ability in relation to life-threatening conditions and diseases such as sepsis, diabetes, cancer, pulmonary and kidney diseases, to name a few. The gold standard method for the measurement of lactate relies on blood sampling, which due to its invasive nature, limits the ability of clinicians in frequent monitoring of patients' lactate levels. Evidence suggests that the optical measurement of lactate holds promise as an alternative to blood sampling. However, achieving this aim requires better understanding of the optical behavior of lactate. The present study investigates the potential deviations of absorbance from the Beer-Lambert law in high concentrations of lactate. To this end, a number of nonlinear models namely support vector machines with quadratic, cubic and quartic kernels and radial basis function kernel are compared with the linear principal component regression and linear support vector machine. Interestingly, it is shown that even in extremely high concentrations of lactate (600 mmol/L) in a phosphate buffer solution, the linear models surpass the performance of the other models.

Machine Learning-Based Presymptomatic Detection of Rice Sheath Blight Using Spectral Profiles

Early detection of plant diseases, prior to symptom development, can allow for targeted and more proactive disease management. The objective of this study was to evaluate the use of near-infrared (NIR) spectroscopy combined with machine learning for early detection of rice sheath blight (ShB), caused by the fungus Rhizoctonia solani. We collected NIR spectra from leaves of ShB-susceptible rice (Oryza sativa L.) cultivar, Lemont, growing in a growth chamber one day following inoculation with R. solani, and prior to the development of any disease symptoms. Support vector machine (SVM) and random forest, two machine learning algorithms, were used to build and evaluate the accuracy of supervised classification-based disease predictive models. Sparse partial least squares discriminant analysis was used to confirm the results. The most accurate model comparing mock-inoculated and inoculated plants was SVM-based and had an overall testing accuracy of 86.1% (N = 72), while when control, mock-inoculated, and inoculated plants were compared the most accurate SVM model had an overall testing accuracy of 73.3% (N = 105). These results suggest that machine learning models could be developed into tools to diagnose infected but asymptomatic plants based on spectral profiles at the early stages of disease development. While testing and validation in field trials are still needed, this technique holds promise for application in the field for disease diagnosis and management.

Application of hyperspectral imaging for spatial prediction of soluble solid content in sweet potato

Visible and near infrared (Vis-NIR) hyperspectral imaging was used for fast detection and visualization of soluble solid content (SSC) in ‘Beijing 553’ and ‘Red Banana’ sweet potatoes. Hyperspectral images were acquired from 420 ROIs of each cultivar of sliced sweet potatoes. There were 8 and 10 outliers removed from ‘Beijing 553’ and ‘Red Banana’ sweet potatoes by Monte Carlo partial least squares (MCPLS). The optimal spectral pretreatments were determined to enhance the performance of the prediction model. Successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS) were employed to select characteristic wavelengths. SSC prediction models were developed using partial least squares regression (PLSR), support vector regression (SVR) and multivariate linear regression (MLR). The more effective prediction performances emerged from the SPA–SVR model with R_p² of 0.8581, RMSEP of 0.2951 and RPD_p of 2.56 for ‘Beijing 553’ sweet potato, and the CARS–MLR model with R_p² of 0.8153, RMSEP of 0.2744 and RPD_p of 2.09 for ‘Red Banana’ sweet potato. Spatial distribution maps of SSC were obtained in a pixel-wise manner using SPA–SVR and CARS–MLR models for quantifying the SSC level in a simple way. The overall results illustrated that Vis-NIR hyperspectral imaging was a powerful tool for spatial prediction of SSC in sweet potatoes.

Visible and near infrared (Vis-NIR) hyperspectral imaging was used for fast detection and visualization of soluble solid content (SSC) in ‘Beijing 553’ and ‘Red Banana’ sweet potatoes.

Development of support vector machine-based model and comparative analysis with artificial neural network for modeling the plant tissue culture procedures: effect of plant growth regulators on somatic embryogenesis of chrysanthemum, as a case study

BACKGROUND

Optimizing the somatic embryogenesis protocol can be considered as the first and foremost step in successful gene transformation studies. However, it is usually difficult to achieve an optimized embryogenesis protocol due to the cost and time-consuming as well as the complexity of this process. Therefore, it is necessary to use a novel computational approach, such as machine learning algorithms for this aim. In the present study, two machine learning algorithms, including Multilayer Perceptron (MLP) as an artificial neural network (ANN) and support vector regression (SVR), were employed to model somatic embryogenesis of chrysanthemum, as a case study, and compare their prediction accuracy.

RESULTS

The results showed that SVR (R2 > 0.92) had better performance accuracy than MLP (R2 > 0.82). Moreover, the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) was also applied for the optimization of the somatic embryogenesis and the results showed that the highest embryogenesis rate (99.09%) and the maximum number of somatic embryos per explant (56.24) can be obtained from a medium containing 9.10 μM 2,4-dichlorophenoxyacetic acid (2,4-D), 4.70 μM kinetin (KIN), and 18.73 μM sodium nitroprusside (SNP). According to our results, SVR-NSGA-II was able to optimize the chrysanthemum's somatic embryogenesis accurately.

CONCLUSIONS

SVR-NSGA-II can be employed as a reliable and applicable computational methodology in future plant tissue culture studies.