Kernel-based Support Vector Machine Research Articles

Retrieval of chromium and mercury concentrations in agricultural soils: Using spectral information, environmental covariates, or a fusion of both?

The universal contamination of arable land with potentially toxic elements (PTEs) poses a severe threat to food security and jeopardizes worldwide efforts to meet the United Nations Sustainable Development Goals (SDGs). How to obtain more reliable concentrations of PTEs in regional agricultural soils is a priority problem to be solved. Multispectral satellite remote sensing, with its advantages of high spatial and temporal resolution, broad coverage, and low cost, offers the potential to acquire spatial distribution of PTEs in agricultural soils over large areas. However, owing to the complexity of soil environments and the insufficiency of spectral information, the mechanism for retrieving concentrations of PTEs in agricultural soils via multispectral satellites is not yet clear, and the accuracy needs to be improved. In this study, we aimed to assess whether employing a fusion of spectral information and environmental covariates, results in more accurate retrievals of PTEs, specifically chromium (Cr) and mercury (Hg), in croplands than does employing spectral information alone. Three machine learning algorithms—kernel-based support vector machine (SVM), neural network-based back propagation neural network (BPNN), and tree-based extreme gradient boosting (XGBoost)—were developed to retrieve Cr and Hg concentrations in agricultural soils. The results showed that the fusion of spectral information and environmental covariates combined with the XGBoost model performed best in retrieving both Cr and Hg concentrations in agricultural soils with coefficient of determination (R2) values of 0.73 and 0.74, respectively. Environmental covariates were important variables for determining Cr and Hg concentrations in agricultural soils, but the ability to retrieve these element concentrations by utilizing spectral information alone was limited. High Cr concentrations occurred in central towns and southern hilly mountains. High Hg concentrations were detected in the northwestern region and southern hilly mountains. The potential of fusing spectral information and environmental covariates to precisely retrieve PTE concentrations in agricultural soils can serve as a reference for agricultural soil health information monitoring worldwide.

Distributionally robust chance-constrained kernel-based support vector machine

Support vector machine (SVM) is a powerful model for supervised learning. This article addresses the nonlinear binary classification problem using kernel-based SVM with uncertainty involved in the input data specified by the first- and second-order moments. To achieve a robust classifier with small probabilities of misclassification, we investigate a distributionally robust chance-constrained kernel-based SVM model. Since the moment information in the original problem becomes unclear/unavailable in the feature space via kernel transformation, we develop a data-driven approach utilizing empirical moments to provide a second-order cone programming (SOCP) reformulation for efficient computation. To speed up the required computations for solving large-size problems in higher dimensional space and/or with more sampling points involved in estimating empirical moments, we further design an alternating direction multipliers-based algorithm for fast computations. Extensive computational results support the effectiveness and efficiency of the proposed model and solution method. Results on public benchmark datasets without any moment information indicate that the proposed approach still works and, surprisingly, outperforms some commonly used state-of-the-art kernel-based SVM models.

Development of new materials for electrothermal metals using data driven and machine learning.

After adopting a combined approach of data-driven methods and machine learning, the prediction of material performance and the optimization of composition design can significantly reduce the development time of materials at a lower cost. In this research, we employed four machine learning algorithms, including linear regression, ridge regression, support vector regression, and backpropagation neural networks, to develop predictive models for the electrical performance data of titanium alloys. Our focus was on two key objectives: resistivity and the temperature coefficient of resistance (TCR). Subsequently, leveraging the results of feature selection, we conducted an analysis to discern the impact of alloying elements on these two electrical properties.The prediction results indicate that for the resistivity data prediction task, the radial basis function kernel-based support vector machine model performs the best, with a correlation coefficient above 0.995 and a percentage error within 2%, demonstrating high predictive capability. For the TCR data prediction task, the best-performing model is a backpropagation neural network with two hidden layers, also with a correlation coefficient above 0.995 and a percentage error within 3%, demonstrating good generalization ability. The feature selection results using random forest and Xgboost indicate that Al and Zr have a significant positive effect on resistivity, while Al, Zr, and V have a significant negative effect on TCR. The conclusion of the composition optimization design suggests that to achieve both high resistivity and TCR, it is recommended to set the Al content in the range of 1.5% to 2% and the Zr content in the range of 2.5% to 3%.

Exploring fetal brain tumor glioblastoma symptom verification with self organizing maps and vulnerability data analysis

Brain tumor glioblastoma is a disease that is caused for a child who has abnormal cells in the brain, which is found using MRI “Magnetic Resonance Imaging” brain image using a powerful magnetic field, radio waves, and a computer to produce detailed images of the body's internal structures it is a standard diagnostic tool for a wide range of medical conditions, from detecting brain and spinal cord injuries to identifying tumors and also in evaluating joint problems. This is treatable, and by enabling the factor for happening, the factor for dissolving the dead tissues. If the brain tumor glioblastoma is untreated, the child will go to death; to avoid this, the child has to treat the brain problem using the scan of MRI images. Using the neural network, brain-related difficulties have to be resolved. It is identified to make the diagnosis of glioblastoma. This research deals with the techniques of max rationalizing and min rationalizing images, and the method of boosted division time attribute extraction has been involved in diagnosing glioblastoma. The process of maximum and min rationalization is used to recognize the Brain tumor glioblastoma in the brain images for treatment efficiency. The image segment is created for image recognition. The method of boosted division time attribute extraction is used in image recognition with the help of MRI for image extraction. The proposed boosted division time attribute extraction method helps to recognize the fetal images and find Brain tumor glioblastoma with feasible accuracy using image rationalization against the brain tumor glioblastoma diagnosis. In addition, 45% of adults are affected by the tumor, 40% of children and 5% are in death situations. To reduce this ratio, in this study, the Brain tumor glioblastoma is identified and segmented to recognize the fetal images and find the Brain tumor glioblastoma diagnosis. Then the tumor grades were analyzed using the efficient method for the imaging MRI with the diagnosis result of partially high. The accuracy of the proposed TAE-PIS system is 98.12% which is higher when compared to other methods like Genetic algorithm, Convolution neural network, fuzzy-based minimum and maximum neural network and kernel-based support vector machine respectively. Experimental results show that the proposed method archives rate of 98.12% accuracy with low response time and compared with the Genetic algorithm (GA), Convolutional Neural Network (CNN), fuzzy-based minimum and maximum neural network (Fuzzy min–max NN), and kernel-based support vector machine. Specifically, the proposed method achieves a substantial improvement of 80.82%, 82.13%, 85.61%, and 87.03% compared to GA, CNN, Fuzzy min–max NN, and kernel-based support vector machine, respectively.

An efficient fake news classification model based on ensemble deep learning techniques

The availability and expansion of social media has made it difficult to distinguish between fake and real news. Information falsification has exponentially increased as a result of how simple it is to spread information through sharing. Social media dependability is also under jeopardy due to the extensive dissemination of false information. Therefore, it has become a research problem to automatically validate information, specifically source, content, and publisher, to identify it as true or false. Despite its limitations, machine learning (ML) has been crucial in the categorization of information. Previous studies suggested three-step methods for categorising false information on social media. In the first step of the process, the data set is subjected to a number of pre-processing processes in order to transform unstructured data sets into structured data sets. The unknowable properties of fake news and the features are extracted by the Lexicon Model in the second stage. In the third stage of this research project, a feature selection method by WOA (Whale Optimization Algorithm) for weight value to tune the classification part. Finally, a Hybrid Classification model that is hybrid with a fuzzy based Convolutional Neural Network and kernel based support vector machine is constructed in order to identify the data pertaining to bogus news. However using single classifier for fake news detection produces the insufficient accuracy. To overcome this issue in this work introduced an improved model for fake news classification. To turn unstructured data sets into structured data sets, a variety of pre-processing operations are used on the data set in the initial phase of the procedure. The unknowable properties of fake news and the features are extracted by the Lexicon Model in the second stage. In the third stage of this research project, a feature selection method by COA (Coati Optimization Algorithm) for weight value to tune the classification part. Finally, an ensemble of RNN (Recurrent Neural Networks), VGG-16 and ResNet50.A classification model was developed to recognise bogus news information. Evaluate each fake news analysis' performance in terms of accuracy, precision, recall, and F1 score. The suggested model, out of all the methodologies taken into consideration in this study, provides the highest outcomes, according to experimental findings

A novel algorithm for classification using a low rank approximation of kernel-based support vector machines with applications

Support vector machines (SVMs), as a powerful technique for classification, are becoming increasingly popular in a wide range of applications. This is simply due to their robustness against several types of model assumptions violations and outliers. The Kernel-based SVM are very useful to capture non-linear patterns in the data, and for classification. However, this kernel-based method could become computationally very challenging because it increases the required time to train data. This increase in computational time is mainly due to the appearance of the kernel in solving the quadratic optimization problem (QOP). In order to tackle this computational complexity, we propose a novel method based on the low-rank approximation, by adapting a truncated Mercer series to the kernels. The quadratic optimization problem in the structure of kernel-based SVM will then be replaced with a much simpler optimization problem. In the proposed approach, the required time for the vector computations and matrix decompositions will be much faster such that these changes lead to efficiently resolve the QOP and ultimately increase efficiency in classification. We finally present some numerical illustrations based on the ROC curves and other classification performance benchmarks considered in this paper to assess the performance of the proposed low-rank approximation to the kernel in SVM structure. The results suggest considerable efficiency improvement has been observed in classification with significant reduction in computational time required to train and forecast the stock market index (S&P 500 index) and promoter recognition in DNA sequences.

Classification of low-functioning and high-functioning autism using task-based EEG signals

Autism is a neurodevelopmental disorder marked by a lack of interpersonal, social, and communication skills, and repetitive and limited behavioral patterns. Autism exhibits a different severity and level of functioning, ranging from high functioning (HF) to low functioning (LF), as per their intellectual/developmental abilities. Identifying the level of the functionality remains crucial in treating and training Autistic children. Brain functioning widely varies during task performance, and the consequent neural representations can act as potential biomarkers. This work aims to classify LF, and HF autistic children based on electroencephalography (EEG), a promising neural underpinning technique acquired while performing cognitive tasks. But task-based EEG data acquisition from autistic children remains challenging and hinders effective classification procedures due to inadequate sample sizes and difficulty acquiring artifact-free signals. Acquired EEG signals were pre-processed for artifact removal to overcome this limitation, and a Gaussian bootstrapping data augmentation technique was employed to strengthen the dataset. Frequency band-specific absolute power, relative power, and derived parameters such as theta-to-alpha ratio and theta-to-beta ratio were extracted. The statistically relevant features were considered for classifying high and low-functioning autism using a kernel-based support vector machine (SVM). Results show that the fine gaussian kernel outperformed the other kernel methods with an average accuracy of 99.9% in the classification process. Thus, this work seems to effectively employ task-based EEG parameters to distinguish the LF and HF groups rather than depending on behavioral measures to categorize autism.

Breast cancer diagnosis using the fast learning network algorithm.

The use of machine learning (ML) and data mining algorithms in the diagnosis of breast cancer (BC) has recently received a lot of attention. The majority of these efforts, however, still require improvement since either they were not statistically evaluated or they were evaluated using insufficient assessment metrics, or both. One of the most recent and effective ML algorithms, fast learning network (FLN), may be seen as a reputable and efficient approach for classifying data; however, it has not been applied to the problem of BC diagnosis. Therefore, this study proposes the FLN algorithm in order to improve the accuracy of the BC diagnosis. The FLN algorithm has the capability to a) eliminate overfitting, b) solve the issues of both binary and multiclass classification, and c) perform like a kernel-based support vector machine with a structure of the neural network. In this study, two BC databases (Wisconsin Breast Cancer Database (WBCD) and Wisconsin Diagnostic Breast Cancer (WDBC)) were used to assess the performance of the FLN algorithm. The results of the experiment demonstrated the great performance of the suggested FLN method, which achieved an average of accuracy 98.37%, precision 95.94%, recall 99.40%, F-measure 97.64%, G-mean 97.65%, MCC 96.44%, and specificity 97.85% using the WBCD, as well as achieved an average of accuracy 96.88%, precision 94.84%, recall 96.81%, F-measure 95.80%, G-mean 95.81%, MCC 93.35%, and specificity 96.96% using the WDBC database. This suggests that the FLN algorithm is a reliable classifier for diagnosing BC and may be useful for resolving other application-related problems in the healthcare sector.

Prediction of Heart Disease using Machine Learning Algorithms with Feature Selection Techniques

Data is an asset in the digital era, and enormous data was generating day by day in all the fields, including the healthcare industry. The data on the healthcare industry data consists of personal information and disease-related information about a patient and stored in various formats and units. Machine learning and Artificial Intelligence techniques will help us analyze the voluminous amount of data to identify the hidden patterns of a specific disease from the healthcare data and help us predict a particular disease in the future. In this paper, we proposed a decision support system to predict heart disease, especially cardiovascular disease, through machine learning algorithms. This system experimented with the reduced set feature of the UCI Machine learning repository dataset using a linear kernel-based support vector machine algorithm. This system has also compared it with other machine learning algorithms such as K-Nearest Neighbours, Decision tree, and Random forest in Python. All four machine learning algorithms' performance has been evaluated based on accuracy, misclassification rate, precision, recall, and f-score value. From the experimental results, SVM with a linear kernel function classification algorithm produces better accuracy of 95.08% compared with others for predicting heart disease.

Automated placental abruption identification using semantic segmentation, quantitative features, SVM, ensemble and multi-path CNN

The placenta is a fundamental organ throughout the pregnancy and the fetus’ health is closely related to its proper function. Because of the importance of the placenta, any suspicious placental conditions require ultrasound image investigation. We propose an automated method for processing fetal ultrasonography images to identify placental abruption using machine learning methods in this paper. The placental imaging characteristics are used as the semantic identifiers of the region of the placenta compared with the amniotic fluid and hard organs. The quantitative feature extraction is applied to the automatically identified placental regions to assign a vector of optical features to each ultrasonographic image. In the first classification step, two methods of kernel-based Support Vector Machine (SVM) and decision tree Ensemble classifier are elaborated and compared for identification of the abruption cases and controls. The Recursive Feature Elimination (RFE) is applied for optimizing the feature vector elements for the best performance of each classifier. In the second step, the deep learning classifiers of multi-path ResNet-50 and Inception-V3 are used in combination with RFE. The resulting performances of the algorithms are compared together to reveal the best classification method for the identification of the abruption status. The best results were achieved for optimized ResNet-50 with an accuracy of 82.88% ± SD 1.42% in the identification of placental abruption on the testing dataset. These results show it is possible to construct an automated analysis method with affordable performance for the detection of placental abruption based on ultrasound images.

Meta-Heuristic Optimized Hybrid Wavelet Features for Arrhythmia Classification

The non-invasive evaluation of the heart through EectroCardioGraphy (ECG) has played a key role in detecting heart disease. The analysis of ECG signals requires years of learning and experience to interpret and extract useful information from them. Thus, a computerized system is needed to classify ECG signals with more accurate results effectively. Abnormal heart rhythms are called arrhythmias and cause sudden cardiac deaths. In this work, a Computerized Abnormal Heart Rhythms Detection (CAHRD) system is developed using ECG signals. It consists of four stages; preprocessing, feature extraction, feature optimization and classifier. At first, Pan and Tompkins algorithm is employed to detect the envelope of Q, R and S waves in the preprocessing stage. It uses a recursive filter to eliminate muscle noise, <i>T</i>-wave interference and baseline wander. As the analysis of ECG signal in the spatial domain does not provide a complete description of the signal, the feature extraction involves using frequency contents obtained from multiple wavelet filters; bi-orthogonal, Symlet and Daubechies at different resolution levels in the feature extraction stage. Then, Black Widow Optimization (BWO) is applied to optimize the hybrid wavelet features in the feature optimization stage. Finally, a kernel based Support Vector Machine (SVM) is employed to classify heartbeats into five classes. In SVM, Radial Basis Function (RBF), polynomial and linear kernels are used. A total of ∼15000 ECG signals are obtained from the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database for performance evaluation of the proposed CAHRD system. Results show that the proposed CAHRD system proved to be a powerful tool for ECG analysis. It correctly classifies five classes of heartbeats with 99.91% accuracy using an RBF kernel with 2^nd level wavelet coefficients. The CAHRD system achieves an improvement of ∼6% over random projections with the ensemble SVM approach and ∼2% over morphological and ECG segment based features with the RBF classifier.

Sensor-Based Gait Analysis for Parkinson’s Disease Prediction

Parkinson’s disease is identified as one of the key neurodegenerative disorders occurring due to the damages present in the central nervous system. The cause of such brain damage seems to be fully explained in many research studies, but the understanding of its functionality remains to be impractical. Specifically, the development of a quantitative disease prediction model has evolved in recent decades. Moreover, accelerometer sensor-based gait analysis is accepted as an important tool for recognizing the walking behavior of the patients during the early prediction and diagnosis of Parkinson’s disease. This type of minimal infrastructure equipment helps in analyzing the Parkinson’s gait properties without affecting the common behavioral patterns during the clinical practices. Therefore, the Accelerometer Sensor-based Parkinson’s Disease Identification System (ASPDIS) is introduced with a kernel-based support vector machine classifier model to make an early prediction of the disease. consequently, the proposed classifier can easily predict various severity levels of Parkinson’s disease from the sensor data. The performance of the proposed classifier is compared against the existing models such as random forest, decision tree, and k-nearest neighbor classifiers respectively. As per the experimental observation, the proposed classifier has more capability to differentiate Parkinson’s from non-Parkinson patients depending upon the severity levels. Also, it is found that the model has outperformed the existing classifiers concerning prediction time and accuracy respectively.

Improved support vector machine classification for imbalanced medical datasets by novel hybrid sampling combining modified mega-trend-diffusion and bagging extreme learning machine model.

To handle imbalanced datasets in machine learning or deep learning models, some studies suggest sampling techniques to generate virtual examples of minority classes to improve the models' prediction accuracy. However, for kernel-based support vector machines (SVM), some sampling methods suggest generating synthetic examples in an original data space rather than in a high-dimensional feature space. This may be ineffective in improving SVM classification for imbalanced datasets. To address this problem, we propose a novel hybrid sampling technique termed modified mega-trend-diffusion-extreme learning machine (MMTD-ELM) to effectively move the SVM decision boundary toward a region of the majority class. By this movement, the prediction of SVM for minority class examples can be improved. The proposed method combines α-cut fuzzy number method for screening representative examples of majority class and MMTD method for creating new examples of the minority class. Furthermore, we construct a bagging ELM model to monitor the similarity between new examples and original data. In this paper, four datasets are used to test the efficiency of the proposed MMTD-ELM method in imbalanced data prediction. Additionally, we deployed two SVM models to compare prediction performance of the proposed MMTD-ELM method with three state-of-the-art sampling techniques in terms of geometric mean (G-mean), F-measure (F1), index of balanced accuracy (IBA) and area under curve (AUC) metrics. Furthermore, paired t-test is used to elucidate whether the suggested method has statistically significant differences from the other sampling techniques in terms of the four evaluation metrics. The experimental results demonstrated that the proposed method achieves the best average values in terms of G-mean, F1, IBA and AUC. Overall, the suggested MMTD-ELM method outperforms these sampling methods for imbalanced datasets.

A kernel-free L1 norm regularized ν-support vector machine model with application

With a view to overcoming a few shortcomings resulting from the kernel-based SVM models, these kernel-free support vector machine (SVM) models are newly promoted and researched. With the aim of deeply enhancing the classification accuracy of present kernel-free quadratic surface support vector machine (QSSVM) models while avoiding computational complexity, an emerging kernel-free ν-fuzzy reduced QSSVM with L1 norm regularization model is proposed. The model has well-developed sparsity to avoid computational complexity and overfitting and has been simplified as these standard linear models on condition that the data points are (nearly) linearly separable. Computational tests are implemented on several public benchmark datasets for the purpose of showing the better performance of the presented model compared with a few known binary classification models. Similarly, the numerical consequences support the more elevated training effectiveness of the presented model in comparison with those of other kernel-free SVM models. What`s more, the presented model is smoothly employed in lung cancer subtype diagnosis with good performance, by using the gene expression RNAseq-based lung cancer subtype (LUAD/LUSC) dataset in the TCGA database.

Estimation of hydraulic conductivity of porous media using data-driven techniques

Abstract Knowledge of hydraulic conductivity (K) is inevitable for sub-surface flow and aquifer studies. Hydrologists and groundwater researchers are employing data-driven techniques to indirectly evaluate K using porous media characteristics as an alternative to direct measurement. The study examines the ability of the Adaptive Neuro-Fuzzy Inference System (ANFIS) to predict the K of porous media using two membership functions (MFs), i.e., triangular and Gaussian, and support vector machine (SVM) via four kernel functions, i.e., linear, quadratic, cubic, and Gaussian. The techniques used easily measurable parameters namely effective and mean grain size, uniformity coefficient, and porosity as input variables. A 70 and 30% dataset is used for the training and testing of models, respectively. The correlation coefficient (R) and root mean square error (RMSE) were used to evaluate the models. The Gaussian MF-based ANFIS model outperformed the triangular model having R and RMSE values of 0.9661 &amp; 0.0010 and 0.9532 &amp; 0.0015, respectively, whereas the quadratic kernel-based SVM model with R and RMSE values of 0.9520 and 0.0015 performs better than the other SVM models. Based on the evaluation of ANFIS and SVM models, the study establishes the efficacy of the Gaussian MF-based ANFIS model in estimating the K of porous media.

Defense against SSDF Attack and PUE Attack in CR-Internet of Vehicles (IoVs) for Millimeter Wave Massive MIMO Beamforming Systems

The Internet of Vehicles (IoV) is witnessed to play the leading role in the future of Intelligent Transportation Systems (ITS). Though many works have focused on IoV improvement, there is still a lack of performance due to insufficient spectrum availability, lower data rates, and the involvement of attackers. This paper considers all three issues by developing a novel mmWave-assisted Cognitive Radio based IoV (CR-IoV) model. The integration of CR in IoV resolves the issue of spectrum management, while mmWave technology ensures symmetry in acquiring higher data rates for Secondary Users (SUs). With the proposed mmWave-assisted CR-IoV model, symmetric improvements in network performance were achieved in three main areas: security, beamforming, and routing. Optimum detection mechanisms isolate malicious Secondary Users (SUs) in the overall network. First, Spectrum Sensing Data Falsification (SSDF) attack is detected by a Hybrid Kernel-based Support Vector Machine (HK-SVM), which is the lightweight Machine Learning (ML) technique. Then, the Primary User Emulation (PUE) attack is detected by a hybrid approach, namely the Fang Algorithm-based Time Difference of Arrival (FA-TDoA) method. Further, security is assured by validating the legitimacy of each SU through a Lightweight ID-based Certificate Validation mechanism. To accomplish this, we employed the Four Q-curve asymmetric cryptographic algorithm. Overall, the proposed dual-step security provisioning approach assures that the network is free from attackers. Next, beamforming is performed for legitimate SUs by a 3D-Beamforming algorithm that relies on Array Factor (AF) and Beampattern Function. Finally, routing is enabled by formulating Forwarding Zone (FZ) based on the forwarding angle. In the forwarding zone, optimal forwarders are selected by the Multi-Objective Whale Optimization (MOWO) algorithm. Here, a new potential score is formulated for fitness evaluation. Finally, the proposed mmWave-assisted CR-IoV model is validated through extensive simulations in the ns-3.26 simulation tool. The evaluation shows better performance in terms of throughput, packet delivery ratio, delay, bit error rate, and detection accuracy.

Targeted Metabolomic Profiling of Plasma Samples in Gastric Cancer by Liquid Chromatography-Mass Spectrometry

Introduction: As the high mortality rate of gastric cancer (GC) is due to delayed diagnosis, early detection is vital for improved patient outcomes. Metabolic deregulation plays an important role in GC. Although various metabolite-level biomarkers for early detection have been assessed, there is still no unified early detection method. We conducted a plasma metabolome study to assess metabolites that may distinguish GC samples from non-GC samples. Methods: Blood samples were collected from 72 GC patients and 29 control participants (non-GC group) at the Tokyo Medical University Hospital between March 2020 and November 2020. Hydrophilic metabolites were identified and quantified using liquid chromatography-time-of-flight mass spectrometry. Differences in metabolite concentrations between the GC and non-GC groups were evaluated using the Mann-Whitney test. The discrimination ability of each metabolite was evaluated by the area under the receiver operating characteristic curve. A radial basis function (RBF) kernel-based support vector machine (SVM) model was developed to assess the discrimination ability of multiple metabolites. The selection of variables used for the SVM utilized a step-wise regression method. Results: Of the 96 quantified metabolites, 8 were significantly different between the GC and non-GC groups. Of these, N1-acetylspermine, succinate, and histidine were used in the RBF-SVM model to discriminate GC samples from non-GC samples. The area under the curve (AUC) of the RBF-SVM model was higher (0.915; 95% CI: 0.865–0.965, p < 0.0001), indicating good performance of the RBF-SVM model. The application of this RBF-SVM to the validation dataset resulted from the AUC of the RBF-SVM model was (0.885; 95% CI: 0.797–0.973, p < 0.0001), indicating the good performance of the RBF-SVM model. The sensitivity of the RBF-SVM model was better (69.0%) than those of the common tumor markers carcinoembryonic antigen (CEA) (10.5%) and carbohydrate antigen 19-9 (CA19-9) (2.86%). The RBF-SVM showed a low correlation with CEA and CA19-9, indicating its independence. Conclusion: We analyzed plasma metabolomics, and a combination of the quantified metabolites showed high sensitivity for the detection of GC. The independence of the RBF-SVM from tumor markers suggested that their complementary use would be helpful for GC screening.

Gray wolf optimization-extreme learning machine approach for diabetic retinopathy detection.

Many works have employed Machine Learning (ML) techniques in the detection of Diabetic Retinopathy (DR), a disease that affects the human eye. However, the accuracy of most DR detection methods still need improvement. Gray Wolf Optimization-Extreme Learning Machine (GWO-ELM) is one of the most popular ML algorithms, and can be considered as an accurate algorithm in the process of classification, but has not been used in solving DR detection. Therefore, this work aims to apply the GWO-ELM classifier and employ one of the most popular features extractions, Histogram of Oriented Gradients-Principal Component Analysis (HOG-PCA), to increase the accuracy of DR detection system. Although the HOG-PCA has been tested in many image processing domains including medical domains, it has not yet been tested in DR. The GWO-ELM can prevent overfitting, solve multi and binary classifications problems, and it performs like a kernel-based Support Vector Machine with a Neural Network structure, whilst the HOG-PCA has the ability to extract the most relevant features with low dimensionality. Therefore, the combination of the GWO-ELM classifier and HOG-PCA features might produce an effective technique for DR classification and features extraction. The proposed GWO-ELM is evaluated based on two different datasets, namely APTOS-2019 and Indian Diabetic Retinopathy Image Dataset (IDRiD), in both binary and multi-class classification. The experiment results have shown an excellent performance of the proposed GWO-ELM model where it achieved an accuracy of 96.21% for multi-class and 99.47% for binary using APTOS-2019 dataset as well as 96.15% for multi-class and 99.04% for binary using IDRiD dataset. This demonstrates that the combination of the GWO-ELM and HOG-PCA is an effective classifier for detecting DR and might be applicable in solving other image data types.

A kernel-free fuzzy reduced quadratic surface [formula omitted]-support vector machine with applications

The kernel-free support vector machine (SVM) models are recently developed and studied to overcome some drawbacks induced by the kernel-based SVM models. To further improve the classification accuracy and computational efficiency of existing kernel-free quadratic surface support vector machine (QSSVM) models, a novel kernel-free ν-fuzzy reduced QSSVM model is proposed. The proposed model utilizes a reduced quadratic surface for nonlinear binary classification as well as reducing the effect of outliers in the data set. Some theoretical properties are rigorously studied, especially, the effects of the parameter ν on the dual feasibility and the number of support vectors. Computational experiments are conducted on some public benchmark data sets to indicate the superior performance of the proposed model over some well-known binary classification models. The numerical results also favors the higher training efficiency of the proposed model over those of other kernel-free SVM models. Moreover, the proposed model is successfully applied to the prodromal detection of Alzheimer’s Disease with good performance, by using the data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database.

Prediction and Classification of COVID-19 Admissions to Intensive Care Units (ICU) Using Weighted Radial Kernel SVM Coupled with Recursive Feature Elimination (RFE).

Healthcare systems have been under immense pressure since the beginning of the COVID-19 pandemic; hence, studies on using machine learning (ML) methods for classifying ICU admissions and resource allocation are urgently needed. We investigated whether ML can propose a useful classification model for predicting the ICU admissions of COVID-19 patients. In this retrospective study, the clinical characteristics and laboratory findings of 100 patients with laboratory-confirmed COVID-19 tests were retrieved between May 2020 and January 2021. Based on patients’ demographic and clinical data, we analyzed the capability of the proposed weighted radial kernel support vector machine (SVM), coupled with (RFE). The proposed method is compared with other reference methods such as linear discriminant analysis (LDA) and kernel-based SVM variants including the linear, polynomial, and radial kernels coupled with REF for predicting ICU admissions of COVID-19 patients. An initial performance assessment indicated that the SVM with weighted radial kernels coupled with REF outperformed the other classification methods in discriminating between ICU and non-ICU admissions in COVID-19 patients. Furthermore, applying the Recursive Feature Elimination (RFE) with weighted radial kernel SVM identified a significant set of variables that can predict and statistically distinguish ICU from non-ICU COVID-19 patients. The patients’ weight, PCR Ct Value, CCL19, INF-β, BLC, INR, PT, PTT, CKMB, HB, platelets, RBC, urea, creatinine and albumin results were found to be the significant predicting features. We believe that weighted radial kernel SVM can be used as an assisting ML approach to guide hospital decision makers in resource allocation and mobilization between intensive care and isolation units. We model the data retrospectively on a selected subset of patient-derived variables based on previous knowledge of ICU admission and this needs to be trained in order to forecast prospectively.