Search

Conventional classification assumes a balanced sample distribution among classes. However, such a premise leads to biased performance over the majority class (with the highest number of instances). The Twin Support Vector Machines (TWSVM) obtained great prominence due to their low computational burden compared to the standard SVM. Besides, traditional machine learning seeks methods whose solution depends on a convex problem or semi-positive definite similarity matrices. Yet, this kind of matrix cannot adequately represent many real-world applications. The above defines the need to use non-negative measures as an indefinite function in a Reproducing Kernel Kreĭn Space (RKKS). This paper proposes a novel approach called Kreĭn Twin Support Vector Machines (KTSVM), which appropriately incorporates indefinite kernels within a TWSVM-based gradient optimization. To code pertinent input patterns within an imbalanced data discrimination, our KTSVM employs an implicit mapping to a RKKS. Also, our approach takes advantage of the TWSVM scheme's benefits by creating two parallel hyperplanes. This promotes the KTSVM optimization in a gradient-descent framework. Results obtained on synthetic and real-world datasets demonstrate that our solution performs better in terms of imbalanced data classification than state-of-the-art techniques. • Kreĭn Twin Support Vector Machines (KTSVM) enhances nonlinear TWSVM with indefinite kernels for imbalanced classification. • KTSVM represents high dimensions with indefinite kernels, improving imbalanced classification. • Our approach transforms the traditional TWSVM's convex dual problem into a gradient-based optimization. • KTSVM builds dual hyperplanes in RKKS, refined with gradient descent. • KTSVM excels over SVM, KSVM, and TWSVM in binary imbalanced classification. [ABSTRACT FROM AUTHOR]

This study examines the impacts of the COVID-19 pandemic on short-term travel behavior and long-term travel preferences among residents of Alabama, using survey data. The study employs a two-step path analysis modeling framework to connect the changes in travel behavior during the pandemic with the expected long-term travel preferences. In the first step, the study develops a model to identify correlates of the travel behavior changes during the pandemic (i.e., reduced trip frequency to grocery/pharmacy, retail/recreation, and work/commuting). In the second step, the study develops a model to establish the association between during-pandemic travel behavior changes and post-pandemic travel preferences. To reduce estimation bias when relying on one single model, the study employs multiple machine learning classifiers such as Random Forest, Adaptive Boosting, Support Vector Machine, K-Nearest Neighbors, and Artificial Neural Network. Average marginal effects are estimated to quantify the correlates of travel impacts due to the pandemic. The results reveal that fear of COVID-19 is significantly associated with reduced travel during the pandemic regardless of the trip type, and people from high-income households will likely travel less after the pandemic. The path analysis connects the correlates of short-term travel impacts of COVID-19 and long-term travel preferences, and identifies variables such as fear of COVID-19 that are significantly linked to reduced travel during the pandemic and also associated with long-term travel preferences. People who reduced their travel during the pandemic are likely to continue traveling less in the future. The study provides valuable insights into the impacts of COVID-19 on travel behavior in Alabama, which can inform the development of local transportation policies. • A statewide travel behavior survey was conducted during the COVID-19 pandemic in Alabama. • The study built models to uncover the impacts of COVID-19 on short-term travel behaviors and long-term travel preferences. • Machine learning integrated into path analysis to explore travel behaviors during COVID-19 effects on long-term preferences. • Estimated marginal effects quantify the impact of the pandemic on short- and long-term travel behavior and preferences. • People in Alabama who reduced their travel during the pandemic are likely to continue traveling less in the future. [ABSTRACT FROM AUTHOR]

• ADHD persisters showed widespread anomalies in microstate and spectral features. • The neural features in remitters exhibited multiple patterns, including normalization, compensation, and fixed abnormalities. • Microstate dynamics and spectral components may be potential markers of distinct ADHD outcomes. We leveraged microstate characteristics and power features to examine temporal and spectral deviations underlying persistent and remittent attention-deficit/hyperactivity disorder (ADHD). 50 young adults with childhood ADHD (28 persisters, 22 remitters) and 28 demographically similar healthy controls (HC) were compared on microstates features and frequency principal components (f-PCs) of eye-closed resting state. Support vector machine model with sequential forward selection (SVM-SFS) was utilized to discriminate three groups. Four microstates and four comparable f-PCs were identified. Compared to HC, ADHD persisters showed prolonged duration in microstate C, elevated power of the delta component (D), and compromised amplitude of the two alpha components (A1 and A2). Remitters showed increased duration and coverage of microstate C, together with decreased activity of D, relatively intact amplitude of A1, and amplitude reduction in A2. The SVM-SFS algorithm achieved an accuracy of 93.59% in classifying persisters, remitters and controls. The most discriminative features selected were those exhibiting group differences. We found widespread anomalies in ADHD persisters in brain dynamics and intrinsic EEG components. Meanwhile, the neural features in remitters exhibited multiple patterns. This study underlines the use of microstate dynamics and spectral components as potential markers of persistent and remittent ADHD. [ABSTRACT FROM AUTHOR]

To develop a magnetic resonance imaging (MRI)-based radiomics model for the preoperative identification of mesenchymal transition (MT) subtype in high-grade serous ovarian cancer (HGSOC). One hundred and eighty-nine patients with histopathologically confirmed HGSOC were enrolled retrospectively. Among the included patients, 55 patients were determined as the MT subtype and the remaining 134 were non-MT subtype. After extracting a total of 204 features from T2-weighted imaging (T2WI) and contrast-enhanced (CE)-T1WI images, the Mann–Whitney U -test, Spearman correlation test, and Boruta algorithm were adopted to select the optimal feature set. Three classifiers, including logistic regression (LR), support vector machine (SVM), and random forest (RF), were trained to develop radiomics models. The performance of established models was evaluated from three aspects: discrimination, calibration, and clinical utility. Seven radiomics features relevant to MT subtypes were selected to build the radiomics models. The model based on the RF algorithm showed the best performance in predicting MT subtype, with areas under the curves (AUCs) of 0.866 (95 % confidence interval [CI]: 0.797–0.936) and 0.852 (95 % CI: 0.736–0.967) in the training and testing cohorts, respectively. The calibration curves, supported with Brier scores, indicated very good consistency between observation and prediction. Decision curve analysis (DCA) showed that the RF-based model could provide more net benefit, which suggested favorable utility in clinical application. The RF-based radiomics model provided accurate identification of MT from the non-MT subtype and may help facilitate personalised management of HGSOC. • Preoperative identification of MT subtype is of importance for patients with HGSOC. • RF-based radiomics model performed best in the prediction of MT subtype. • RF-based radiomics model had favorable calibration performance and clinical utility. [ABSTRACT FROM AUTHOR]

RNA modification, a posttranscriptional regulatory mechanism, significantly influences RNA biogenesis and function. The accurate identification of modification sites is paramount for investigating their biological implications. Methods for encoding RNA sequence into numerical data play a crucial role in developing robust models for predicting modification sites. However, existing techniques suffer from limitations, including inadequate information representation, challenges in effectively integrating positional and sequential information, and the generation of irrelevant or redundant features when combining multiple approaches. These deficiencies hinder the effectiveness of machine learning models in addressing the performance challenges associated with predicting RNA modification sites. Here, we introduce a novel RNA sequence feature representation method, named BiPSTP, which utilizes bidirectional trinucleotide position-specific propensities. We employ the parameter ξ to denote the interval between the current nucleotide and its adjacent forward or backward dinucleotide, enabling the extraction of positional and sequential information from RNA sequences. Leveraging the BiPSTP method, we have developed the prediction model mRNAPred using support vector machine classifier to identify multiple types of RNA modification sites. We evaluate the performance of our BiPSTP method and mRNAPred model across 12 distinct RNA modification types. Our experimental results demonstrate the superiority of the mRNAPred model compared to state-of-art models in the domain of RNA modification sites identification. Importantly, our BiPSTP method enhances the robustness and generalization performance of prediction models. Notably, it can be applied to feature extraction from DNA sequences to predict other biological modification sites. [ABSTRACT FROM AUTHOR]

• Introduce a novel "yielding dilemma zone" model to investigate driver behaviour at nonsignalised crosswalks. • Provides valuable insights into driver behaviour and its impact on pedestrian safety By employing SVM, ANN, and BLR. • Suggest potential real-world applications to enhance pedestrian safety and traffic management at non-signalised crosswalks. In developing countries like China, non-signalised crosswalks represent a considerable risk to pedestrian safety. This study introduces a "yielding dilemma zone" model to tackle the decision-making challenges encountered by drivers at these crosswalks. Field data collected via drones at three non-signalised crosswalks in Xi'an (China) facilitated the analysis of 1687 instances of driver yielding behaviour. The study employed Support Vector Machine (SVM), Artificial Neural Network (ANN), and Binary Logistic Regression (BLR) models to establish the boundaries of the dilemma zone. Additionally, it utilized surrogate safety metrics, such as the deceleration rate required to stop (DRS) and time to crossing (TC), to compare risks between dilemma and non-dilemma zones. Each modelling approach provided insightful findings. The SVM and ANN models identified the speed-distance dilemma zones, while the BLR model determined the multifactorial dilemma zones by considering characteristics such as pedestrian, vehicle, and traffic condition. The findings underscore the significance of the "yielding dilemma zone" for understanding driver-pedestrian interactions at non-signalised crosswalks. The innovative multi-model approach uncovered that pedestrian speed, vehicle type, vehicle approach speed, adjacent vehicle number, driving directions, and pedestrian crossing stage significantly influence this interaction. indicated a higher probability of collisions within dilemma zones compared to non-dilemma zones across all models. The study suggests that the SVM model could be instrumental in creating passive physical warning signs, while the ANN and BLR models could significantly contribute to real-time pedestrian and driver assistance systems. The "yielding dilemma zone" model, with its comprehensive approach, presents a promising avenue for enhancing pedestrian safety, intelligent transportation systems, and traffic management at non-signalised crosswalks. [ABSTRACT FROM AUTHOR]

The placenta is a membrane that separates the fetus from the maternal circulation, and in addition to protecting the fetus, plays a key role in fetal growth and development. With increasing drug use in pregnancy, it is imperative that reliable models of estimating placental permeability and safety be established. In vitro methods and animal models such as rodent placenta are limited in application since the species-specific nature of the placenta prevents meaningful extrapolations to humans. In this regard, in silico approaches such as quantitative structure-property relationships (QSPRs) are useful alternatives. However, despite evidence that drug transport across the placenta is stereoselective (i.e., governed by the spatial arrangement of the atoms in a molecule), many QSPR models for placental transfer have been built using 2D descriptors that do not account for chirality and stereochemistry. In this study, we apply a chirality-sensitive and proven QSPR methodology titled "EigenValue ANalySis" (EVANS) to build QSPR models for placental transfer. We deploy EVANS along with robust machine learning algorithms to build (i) regression models on a dataset of environmental chemicals (dataset PD I) followed by (ii) classification models on a set of drug-like compounds (dataset PD II). The best models were found to achieve state-of-the-art performance, with the support vector machine algorithm returning r train 2 = 0.85 , r test 2 = 0.75 for PD I, and the logistic regression algorithm giving accuracy 0.88 and F1 score 0.93 for PD II. The best models were interpreted with the Shapley Additive Explanations paradigm, and it was found that autocorrelation descriptors are crucial for modelling placental permeability. In conclusion, we demonstrate the need of a chirality-sensitive approach for modelling placental transfer of chemicals, and present two predictive QSPR models that may reliably be used for prediction of placental transfer. • A chirality-sensitive methodology titled "EigenValue ANalySis" (EVANS) is applied to build QSPR models for placental transfer • SVM regression models on a dataset of environmental chemicals returned r train 2 = 0.85 , r test 2 = 0.75 • Logistic regression models on a set of drug-like compounds had accuracy 0.88 and F1 score 0.93 • We demonstrate the need of a chirality-sensitive approach for modelling placental transfer of chemicals [ABSTRACT FROM AUTHOR]

In recent years, machine learning (ML) algorithms have gained substantial recognition for ecological modeling across various temporal and spatial scales. However, little evaluation has been conducted for the prediction of soil organic carbon (SOC) on small data sets commonly inherent to long-term soil ecological research. In this context, the performance of ML algorithms for SOC prediction has never been tested against traditional process-based modeling approaches. Here, we compare ML algorithms, calibrated and uncalibrated process-based models as well as multiple ensembles on their performance in predicting SOC using data from five long-term experimental sites (comprising 256 independent data points) in Austria. Using all available data, the ML-based approaches using Random forest and Support vector machines with a polynomial kernel were superior to all process-based models. However, the ML algorithms performed similar or worse when the number of training samples was reduced or when a leave-one-site-out cross validation was applied. This emphasizes that the performance of ML algorithms is strongly dependent on the data-size related quality of learning information following the well-known curse of dimensionality phenomenon, while the accuracy of process-based models significantly relies on proper calibration and combination of different modeling approaches. Our study thus suggests a superiority of ML-based SOC prediction at scales where larger datasets are available, while process-based models are superior tools when targeting the exploration of underlying biophysical and biochemical mechanisms of SOC dynamics in soils. Therefore, we recommend applying ensembles of ML algorithms with process-based models to combine advantages inherent to both approaches. • We evaluated process-based (PB) and machine-learning (ML) models for SOC prediction. • Random forest & support vector machine predicted SOC superior to PB models. • Reduced data availability and quality significantly weakened ML model performance. • ML-assisted PB model ensembles strongly improved model performance. [ABSTRACT FROM AUTHOR]

Underground hydrogen storage (UHS) offers a promising approach for the storage of significant volumes of hydrogen gas (H 2) within deep geological formations, which can later be utilized for energy generation when necessary. Interfacial tension (IFT) between H 2 and the formation brine plays a vital role in influencing the distribution of H 2 at the pore scale and, ultimately, the storage capacity. In this research, we developed four intelligent models: Decision Trees (DT), Random Forests (RF), Support Vector Machines (SVM), and Multi-Layer Perceptron (MLP). These models were designed to predict the IFT utilizing pressure, temperature, and molality. Additionally, we fine-tuned three explicit correlations previously developed in our research. To assess the influence of each parameter on IFT, we conducted a comprehensive analysis of raw data to exclude doubtful samples. This was followed by rigorous model development, including hyperparameter tuning, and finally, an examination of developed models using testing data. The results clearly demonstrate the superiority of the RF model, achieving high accuracy and reliability with coefficients of determination (R2), root mean square error (RMSE), and average absolute relative deviation (AARD) values of 0.96, 1.50, and 1.84 %, respectively. The exemplary performance of the RF model can be attributed to its inherent characteristics. The ensemble approach of RF excels in capturing complex data relationships, thereby enhancing predictive accuracy and solidifying its superiority over other models in this study. Furthermore, the feature importance analysis revealed that temperature has the most significant influence, followed by molality and pressure. Moreover, we assessed the predictive accuracy of these models through external testing using data not used in the initial training and testing stages. Our study highlights the exceptional predictive power of the RF model, emphasizing the practical importance of selecting RF for enhanced accuracy and reliability. The proposed method shows significant potential for industrial applications, especially in optimizing underground H 2 storage. • Various ML techniques were employed to calculate the IFT in H 2 -brine systems. • Different statistical metrics were employed to assess the models' performance. • The developed ML approaches showed acceptable predictions of the IFT values. • The RF based model outperformed other intelligent models. • The reliability and accuracy of models were demonstrated using external testing. [ABSTRACT FROM AUTHOR]

Support vector machine is one of the most used machine learning algorithms with a comprehensive mathematical infrastructure. The power behind the algorithm is the kernel trick that enables the model to overcome non-linear data distributions by using functions that satisfy the Mercer condition. Undoubtedly, the radial basis function (RBF) is among the most widely used of these functions. The RBF kernel, which has a Gaussian curve, performs local boundary surfaces and supports the generalization capability of the model. In this study, a novel kernel function named Logarithmic Kernel Function (LKF), which has a Gaussian-like curve is presented. The crucial contribution of the LKF is that it can model the dataset better than other similarly shaped kernels in the case of a few training samples (10% and 30%). In the study, 6 classifications and 5 regression sets are handled to compare kernels. Instead of giving the hyperparameters needed by the models manually, they are estimated through the Tree Parzen Estimator, which is based on Sequential Modeling Optimization. This estimation process is repeated 10 times and the average results are obtained. The proposed LKF surpasses the most competitive kernels in various classification and regression tasks. • The novel Gaussian-like Logarithmic Kernel Function (LKF) was proposed. • LKF has been shown to work better than other kernels of similar shape at a small amount of training samples. • For a fair comparison, the parameters of kernels were not determined manually. • Tree Parzen Estimator optimization methods was preferred for parameter selection. • In comparisons made with 6 classification and 6 regression tasks. [ABSTRACT FROM AUTHOR]

Online monitoring of social networks offers exciting features for platforms, enabling both technical and behavioral analysis. Numerous studies have explored the adaptation of traditional quality control methods for detecting change points within social networks. However, the current research studies face limitations such as an overreliance on case-based attributes, high computational costs, poor scalability with large networks, and low sensitivity in fast change point detection. This paper proposes a novel algorithm for social network monitoring using One-Class Support Vector Machines (OC-SVMs) to address these limitations. Additionally, using both nodal and network-level attributes makes it versatile for diverse social network applications and effectively detecting network disturbances. The algorithm utilizes a well-defined training data dictionary with an updating procedure for evolutionary networks, enhancing memory and time efficiency by reducing the processing of input data. Extensive numerical experiments are conducted using an EpiCNet model to simulate interactions in an online social network, covering six change scenarios to evaluate the proposed methodology. The results show lower Average Run Length (ARL) and Expected Delay Detection (EDD), demonstrating the superior accuracy and effectiveness of the OC-SVM algorithm compared to alternative methods. Applying OC-SVM to the Enron Email network indicates its capability to identify change points, reflecting the tumultuous timeline that led to Enron's downfall. This further validates the substantial advancement of OC-SVM in social network monitoring and opens doors to broader real-world applications. [ABSTRACT FROM AUTHOR]

Radial fans play a critical role as indispensable turbomachines in various industrial sectors. However, the conventional manufacturing process for these fans is often characterized by its resource-intensive and time-consuming nature. Traditionally, computational fluid dynamics (CFD) has been the go-to method for predicting and analyzing the performance of radial fans at different geometrical and operational conditions. Yet, in recent years, the rapid advancements in machine learning (ML) and deep learning (DL) techniques, particularly the rise of artificial neural networks (ANNs), have propelled significant progress in the field of predicting and optimizing the performance of radial fans. The present study aims to analyze the performance of a radial fan through a comprehensive experimental investigation and a meticulous three-dimensional numerical simulation. Subsequently, in order to predict the off-design performance of the fan, an extensive set of numerical simulations is conducted at various volumetric flow rates and rotational speeds. These simulations are used to analyze the fan performance and identify the most efficient operating condition. Moreover, the simulations serve as inputs for a finely-tuned ANN architecture. The predictive accuracy of the ANN model for both interpolation and extrapolation cases is then compared against two alternative techniques, namely support vector machine (SVM) and random forest (RF). The results explicitly highlight the superiority of the ANN model in terms of its predictive accuracy, thereby solidifying its position as the most reliable method for predicting the performance of radial fans. [ABSTRACT FROM AUTHOR]

The role of surgery in spontaneous intracerebral hemorrhage (SICH) remains controversial. We aimed to use explainable machine learning (ML) combined with propensity-score matching to investigate the effects of surgery and identify subgroups of patients with SICH who may benefit from surgery in an interpretable fashion. We conducted a retrospective study of a cohort of 282 patients aged ≥21 years with SICH. ML models were developed to separately predict for surgery and surgical evacuation. SHapley Additive exPlanations (SHAP) values were calculated to interpret the predictions made by ML models. Propensity-score matching was performed to estimate the effect of surgery and surgical evacuation on 90-day poor functional outcomes (PFO). Ninety-two patients (32.6%) underwent surgery, and 57 patients (20.2%) underwent surgical evacuation. A total of 177 patients (62.8%) had 90-day PFO. The support vector machine achieved a c-statistic of 0.915 when predicting 90-day PFO for patients who underwent surgery and a c-statistic of 0.981 for patients who underwent surgical evacuation. The SHAP scores for the top 5 features were Glasgow Coma Scale score (0.367), age (0.214), volume of hematoma (0.258), location of hematoma (0.195), and ventricular extension (0.164). Surgery, but not surgical evacuation of the hematoma, was significantly associated with improved mortality at 90-day follow-up (odds ratio, 0.26; 95% confidence interval, 0.10–0.67; P = 0.006). Explainable ML approaches could elucidate how ML models predict outcomes in SICH and identify subgroups of patients who respond to surgery. Future research in SICH should focus on an explainable ML-based approach that can identify subgroups of patients who may benefit functionally from surgical intervention. [ABSTRACT FROM AUTHOR]

Identifying valuable mineral resources is obligatory for human development and survival due to their scientific and economic implications. Hyperspectral Remote Sensing (HRS) is an efficient technique that empowers us to precisely identify and map altered minerals based on spectral absorption curves in the VNIR and SWIR range of the electromagnetic spectrum. Machine Learning Algorithms (MLAs) are a subclass of Artificial Intelligence (AI) that improves and automate HRS based lithological mapping using spectra based classification approach. The present study evaluates various MLAs in identifying and mapping hydrothermally altered and weathered minerals such as kaolinite, talc, kaosmec, and montmorillonite. The study was performed with the Airborne Visible Infrared Imaging Spectrometer- Next Generation (AVIRIS-NG) hyperspectral dataset of Jahazpur town of Bhilwara district in the state of Rajasthan, India (75° 9′ 57′' E, 25° 32′ 24″ N). The Spectral Angle Mapper (SAM) algorithm was applied to create a reference mineral distribution map for the target mineral classes. Further, the reference map has been verified with the field validation survey. A total of 828 pixels were identified as end member pixels. The obtained endmember pixels were further used for developing predictive machine learning based models for mapping the mineral resources using eight supervised classifiers, namely Support Vector Machine, k-Nearest Neighbor, Decision Tree, Random Forest, Logistic Regression, Artificial Neural Network, Linear Discriminant Analysis, and Naïve Bayes. The performance of the classifiers was measured using the kappa coefficient, overall accuracy, precision, recall, F1-score, Matthews's correlation coefficient, and ROC curve. It was observed in the study that the Support Vector Machine outperformed all the classifiers in terms of overall accuracy and AVIRIS-NG data poses an excellent potential for mineral mapping using machine learning based models in a diverse mountainous area. [ABSTRACT FROM AUTHOR]

The identification of crop diversity in today's world is very crucial to ensure adaptation of the crop with changing climate for better productivity as well as food security. Towards this, Hyperspectral Remote Sensing (HRS) is an efficient technique that offers the opportunity to discriminate crop types based on morphological as well as physiological features due to availability of contiguous spectral bands. The current work utilized the benefits of Airborne Visible Infrared Imaging Spectrometer- New Generation (AVIRIS-NG) data and explored the techniques for classification and identification of crop types. The endmembers were identified using the Geo-Stat Endmember Extraction (GSEE) algorithm for pure pixels identification and to generate the spectral library of the different crop types. Spectral feature comparison was done among AVIRIS-NG, Analytical Spectral Device (ASD)-Spectroradiometer and Continuum Removed (CR) spectra. The best-fit spectra obtained with the Reference ASD-Spectroradiometer and Pure Pixel spectral library were then used for crop discrimination using the ten supervised classifiers namely Spectral Angle Mapper (SAM), Spectral Information Divergence (SID), Support Vector Machine (SVM), Minimum Distance Classifier (MDC), Binary Encoding, deep learning-based Convolution Neural Network (CNN) and different algorithms of Ensemble learning such as Tree Bag, AdaBoost (Adaptive Boosting), Discriminant and RUSBoost (Random Under Sampling). In total, nine crop types were identified, namely, wheat, maize, tobacco, sorghum, linseed, castor, pigeon pea, fennel and chickpea. The performance evaluation of the classifiers was made using various metrics like Overall Accuracy, Kappa Coefficient, Precision, Recall and F1 score. The classifier 2D-CNN was found to be the best with Overall Accuracy, Kappa Coefficient, Precision, Recall and F1 score values of 89.065 %, 0.871, 87.565%, 89.541% and 88.678% respectively. The output of this work can be utilized for large scale mapping of crop types at the species level in a short interval of time with high accuracy. [ABSTRACT FROM AUTHOR]

Crop mapping using remote sensing data is essential for predicting crop yield and future requirements. However, it can be accomplished by having full ground truth data corresponding to the samples captured by remote sensors. In a practical scenario, generally, we have limited ground-truth data where available crop samples are less than the number of spectral bands of a hyperspectral image. This paper proposes a sample selection method to select samples from a hyperspectral image based on similarity measure and fuzziness in the samples. We use a single-ground truth pixel of each crop for crop-sample selection. Firstly, we select the number of samples for each crop class using Spectral Information Diversion and tangent of Spectral Angle Mapper ( SS tan ) measure from a selected region of the hyperspectral image. Then, we compute fuzziness for each selected crop sample to generate a set of final samples. We have evaluated the proposed method on an Airborne Visible near InfraRed Imaging Spectrometer-Next Generation (AVIRIS-NG) dataset and two benchmark datasets, i.e., Indian Pines and Salinas. Finally, classification results using three supervised classifiers, namely, support vector machine, k-nearest neighbourhood and random forest, have been shown in terms of overall accuracy and kappa coefficient. It can be demonstrated that the proposed method outperforms some prevailing methods by at least 4.5 % for 5 samples and 11% for 10 samples for the Indian Pines dataset. [ABSTRACT FROM AUTHOR]

In the present study, we have integrated the spectral anomaly map derived from airborne visible-infrared imaging spectrometer-next generation (AVIRIS-NG) data and residual ground geophysical anomaly maps (gravity and magnetic) for identifying the mafic cumulates and chromitites in the Sittampundi Layered-Complex (SLC), Tamil Nadu, India. A spectral map of chromite bearing mafic cumulates was prepared by implementing the matched filtering (MF) algorithm on the selected spectral bands of AVIRIS-NG data using the mixed image spectra of mafic cumulate and chromitite as an end member. A low pass filter was applied in the spectral map to extract the clusters of coherently distributed spectrally anomalous pixels representing the surface exposure of mafic cumulates. Concurrently, the ground gravity and magnetic data were processed for deriving the residual gravity and magnetic anomaly maps. Subsequently, an integration of spectral map, residual gravity, and magnetic anomaly maps was attempted using the support vector machine (SVM) algorithm to identify the potential sites of chromitite occurrence. Results were verified using the geological map and the field data on surface exposures of mafic cumulates. This research demonstrates that integration of remote sensing data and geophysical anomaly maps has immense potential to detect the surface distribution of chromite bearing mafic cumulates in the layered complex and the same methods can be replicated for carrying out mineral prospecting in the other parts of the world. [ABSTRACT FROM AUTHOR]

The transition period in dairy cows is a critical stage and peripartum oxidative status, negative energy balance (NEB), and inflammation are highly prevalent. Fecal microbial metabolism is closely associated with blood oxidative status and nonesterified fatty acids (NEFA) levels. Here, we investigated dynamic changes in total oxidative status markers and NEFA in blood, fecal microbiome, and metabolome of 30 dairy cows during transition (−21, −7, +7, +21 d relative to calving). Then the Bayesian network and 9 machine-learning algorithms were applied to dismantle their relationship. Our results show that the oxidative status indicator (OSI) of −21, −7, +7 d was higher than +21 d. The plasma concentration of NEFA peaked on +7 d. For fecal microenvironment, a decline in bacterial α diversity was observed at postpartum and in bacterial interactions at +7 d. Conversely, microbial metabolites involved in carbohydrate, lipid, and energy metabolism increased on +7 d. A correlation analysis revealed that 11 and 10 microbial metabolites contributed to OSI and NEFA variations, respectively (arc strength >0.5). The support vector machine (SVM) radial model showed the highest average predictive accuracy (100% and 88.9% in the test and external data sets) for OSI using 1 metabolite and 3 microbiota. The SVM radial model also showed the highest average diagnostic accuracy (100% and 91% in the test and external data sets) for NEFA with 2 metabolites and 3 microbiota. Our results reveal a relationship between variation in the fecal microenvironment and indicators of oxidative status, NEB, and inflammation, which provide a theoretical basis for the prevention and precise regulation of peripartum oxidative status and NEB. [ABSTRACT FROM AUTHOR]

• The wheat traits and nitrogen status retrieved from the Landsat-7 images. • Three algorithms are compared to predict wheat parameters and nitrogen status. • The ANN and SVM algorithms can effectively predict plant nitrogen percentage. • The DNN algorithm can predict the wheat parameters better than SVM and ANN. Monitoring wheat biophysical variables during the growing season is essential for achieving precision agriculture's important goals. This study estimated plant height, leaf development stage, leaf area index, dry matter, and plant nitrogen content in wheat from Landsat-7 images in North Iran using machine learning algorithms. Forty points were selected from ten wheat fields (270 ha), and the agronomic parameters were recorded during the 2011–2012 growing season. The machine learning models applied in this study were Artificial Neural Network (ANN), Support Vector Machine (SVM), and Deep Neural Network (DNN). Results suggest that the DNN model was the best for predicting all the parameters, except for the nitrogen percentage, in which SVM and DNN have the same accuracy. The DNN model R2 was 0.82, 0.95, 0.95, and 0.90 RMSE (Root Mean Square Error) 9.61, 0.46, 0.47 and 1.2 and MAE (Mean Absolute Error) 26.4, 0.34, 0.28, and 0.84 for estimating plant height, leaf area index, leaf development stage and dry matter biomass, respectively. The DNN and SVM algorithms can effectively predict plant nitrogen percentage with identical R2, RMSE, and MAE values of 0.83–0.84, 0.44–0.46, and 0.33–0.36, respectively. These results show that machine learning models have great potential to provide timely and accurate data for better wheat management under field conditions from satellite imagery. [ABSTRACT FROM AUTHOR]

The aerosol particles present in the atmospheric region mainly affect the climate radiative forcing directly by scattering & absorbing the sunlight. Also, it indirectly influences the formation of clouds, precipitation and acts as a considerable uncertainty in assessing Earth's radiation budget. Determination of aerosol type is significant in characterizing the aerosol role in the atmospheric processes, feedback, and climate models. This paper proposes two aerosol classification models, one based on the source and another based on the composition, to classify the aerosols using aerosol optical properties. The source based aerosol classification method helps to identify the sources which cause pollution in a particular region. Based on the results, proper control measures can be taken to reduce pollution. The composition based aerosol classification helps to identify the nature of aerosol types, such as absorbing or non-absorbing. This classification helps to study the climate of the Kanpur region. The aerosol data is taken from AERONET (AErosol RObotic NETwork) for the period 2002–2018 for the Kanpur region. The composition based aerosol classification model uses Single Scattering Albedo (SSA), Angstrom Exponent (AE), and Fine Mode Fraction (FMF) parameters to categorize aerosols based on their composition. The source based aerosol classification model classifies the aerosols based on values of AE and Aerosol Optical Depth (AOD) and describes the source of the aerosol particles. Knowledge of aerosol sources and compositions helps execute policies or controls to reduce aerosol concentrations. Machine learning algorithms, Naïve Bayes, K Nearest Neighbor, Decision Tree, Support Vector Machine, and Random Forest are used to validate classification schemes. The performance analysis of machine learning algorithms is compared using ten different metrics, and the results are also compared with the existing aerosol classification models. The results of the classification show that the source based aerosols of the desert and arid background and the composition based aerosols of types, Mixture Absorbing, Coarse absorbing (Dust), and Black Carbon are dominant over the Kanpur region during the study period considered. The Number of non-absorbing (scattering) type aerosols are least in the study region considered during the study period at all the seasons. It is found that the Random Forest and Decision Tree models outperform the other machine learning models considered and the existing classification models in terms of accuracy (99.55 %) and other performance metrics considered. [ABSTRACT FROM AUTHOR]

To investigate the value of non-contrast CT (NCCT)-based two-dimensional (2D) radiomics features in predicting haematoma expansion (HE) after spontaneous intracerebral haemorrhage (ICH) and compare its predictive ability with the three-dimensional (3D) signature. Three hundred and seven ICH patients who received baseline NCCT within 6 h of ictus from two stroke centres were analysed retrospectively. 2D and 3D radiomics features were extracted in the manner of one-to-one correspondence. The 2D and 3D models were generated by four different machine-learning algorithms (regularised L1 logistic regression, decision tree, support vector machine and AdaBoost), and the receiver operating characteristic (ROC) curve was used to compare their predictive performance. A robustness analysis was performed according to baseline haematoma volume. Each feature type of 2D and 3D modalities used for subsequent analyses had excellent consistency (mean ICC >0.9). Among the different machine-learning algorithms, pairwise comparison showed no significant difference in both the training (mean area under the ROC curve [AUC] 0.858 versus 0.802, all p> 0.05) and validation datasets (mean AUC 0.725 versus 0.678, all p> 0.05), and the 10-fold cross-validation evaluation yielded similar results. The AUCs of the 2D and 3D models were comparable either in the binary or tertile volume analysis (all p> 0.5). NCCT-derived 2D radiomics features exhibited acceptable and similar performance to the 3D features in predicting HE, and this comparability seemed unaffected by initial haematoma volume. The 2D signature may be preferred in future HE-related radiomic works given its compatibility with emergency condition of ICH. • The 2D mode for lesion segmentation simplifies radiomic analysis process in ICH. • The 2D signature has the potential to reduce errors caused by manual delineation. • The 2D model possesses similar performance to the 3D model in predicting HE. • This comparability seems not affected by the baseline hematoma volume. [ABSTRACT FROM AUTHOR]

To develop and validate a non-invasive computed tomography (CT)-based radiomics model for predicting vasculogenic mimicry (VM) status in lung adenocarcinoma (LA). Two hundred and three patients with LA were enrolled retrospectively and grouped into training and test groups with a ratio of 7:3. Uni- and multivariate logistic regression analyses were performed in the training cohort to screen the independent clinical and radiological factors for VM, and the clinical model was then established. A radiomics model was established based on the rad-scores through support vector machine (SVM). A radiomics nomogram model was subsequently constructed by combining the rad-score with clinical–radiological factors. The receiver operating characteristic curve (ROC), calibration curves, and decision curve analysis (DCA) were conducted to evaluate the performance of the three models. Nine selected radiomics features were selected for the radiomics model and the maximum length and spiculation sign were constructed for the clinical model. The radiomics nomogram model integrating the maximum length, spiculation sign, and rad-score yielded the best AUC in both the training (AUC = 0.925) and test cohorts (AUC = 0.978), in comparison with the radiomics model (AUC = 0.907 and 0.964, in both the training and test cohorts) and the clinical model (AUC = 0.834 and 0.836 in both training and test cohorts). The CT-based radiomics nomogram model showed satisfying discriminating performance for preoperatively and non-invasively predicting VM expression status in LA patients. • Nine optimal radiomics features are correlated with VM expression of LA. • The maximum length, spiculation sign and Rad-score are risk factors of VM + LA. • The radiomics nomogram model shows best discriminating capability for VM expression. [ABSTRACT FROM AUTHOR]

Using a machine learning algorithm, individuals can be accurately identified from their muscle activation patterns during gait, leading to the concept of individual muscle activation signatures. Are muscle activation signatures robust across different walking speeds? We used an open dataset containing electromyographic (EMG) signals from 8 lower limb muscles in 50 asymptomatic adults walking at 5 speeds (extremely slow, very slow, slow, spontaneous, and fast). A machine learning approach classified the EMG profiles based on similar (intra-speed classification) or different (inter-speed classification) walking speeds as training and testing conditions. Intra-speed median classification rates of muscle activation profiles increased with walking speed, from 92 % for extremely slow, to 100 % for self-selected fast walking conditions. Inter-speed median classification rates increased when the speed of the training condition was closer to that of the testing condition. Higher median classification rates were found across slow, spontaneous, and fast walking speed conditions, from 56 % to 96 %, compared with classification rates involving extremely and very slow walking speed conditions, from 6 % to 62 %. Our findings reveal that i) muscle activation signatures are detectable for a large range of walking speeds, even those involving different gait strategies (intra-speed median classification rates from 92 % to 100 %), and ii) muscle activation signatures observed during very low walking speeds are not consistent with those observed at higher speeds, suggesting a difference in motor control strategy. Caution should therefore be exercised when assessing gait deviations of a slow walking patient against a normative database obtained at higher speed. Identifying the robustness of individual muscle activation signatures across different movements could help in detecting changes in motor control, otherwise difficult to detect on classical time-varying EMG patterns. • Signatures are distinctive patterns by which an individual can be identified. • Individual muscle activation signatures are revealed using EMG patterns during gait. • Muscle activation signatures are detectable for a large range of walking speeds. • Muscle activation signatures are consistent across low to high walking speeds. • A change in motor control occurs around the walking speed threshold of 0.6 m/s. [ABSTRACT FROM AUTHOR]

Our study aimed to explore the abnormal spontaneous brain activity by regional homogeneity (ReHo) and its association with cognitive function to understand the neuropathology of major depressive disorder (MDD). ReHo was used to investigate brain activities of 60 patients with first-episode drug-naive MDD and 60 healthy controls (HCs). Partial correlation analysis was conducted on altered ReHo values and the severity of symptoms and cognitive deficits. Moreover, support vector machine analysis was used to evaluate the accuracy of abnormal ReHo values in distinguishing patients with MDD from HCs. Compared with HCs, patients with MDD showed significantly increased ReHo values in the right cerebellum crus2 and right thalamus and decreased ReHo values in the right angular gyrus (AG) and right precuneus (PCUN). The ReHo values in right cerebellum crus2 and right AG were positively associated with working memory and visual learning, respectively. Furthermore, the combination of ReHo values in the right cerebellum crus2 and right PCUN discriminated the patients with MDD from HCs with specificity, sensitivity, and accuracy of 0.9688, 0.6250, and 0.90, respectively. The design of repeated cross-sectional surveys does not allow analyses of within individual changes. Our study revealed that the pathophysiology mechanism of cognitive deficits in MDD may be related to abnormal spontaneous brain activity. Moreover, the combination of ReHo values in the right cerebellum crus2 and right PCUN can be used to discriminate patients with MDD from HCs effectively. • Widespread ReHo alterations and cognitive dysfunction in MDD • Cognitive deficits in MDD were related to abnormal spontaneous brain activity. • Abnormalities might distinguish patients with MDD from healthy controls. [ABSTRACT FROM AUTHOR]

10-Hz repetitive transcranial magnetic stimulation(rTMS) and intermittent theta-burst stimulation(iTBS) over left prefrontal cortex are FDA-approved, effective options for treatment-resistant depression (TRD). Optimal prediction models for iTBS and rTMS remain elusive. Therefore, our primary objective was to compare prediction accuracy between classification by frontal theta activity alone and machine learning(ML) models by linear and non-linear frontal signals. The second objective was to study an optimal ML model for predicting responses to rTMS and iTBS. Two rTMS and iTBS datasets (n = 163) were used: one randomized controlled trial dataset (RCTD; n = 96) and one outpatient dataset (OPD; n = 67). Frontal theta and non-linear EEG features that reflect trend, stability, and complexity were extracted. Pretreatment frontal EEG and ML algorithms, including classical support vector machine(SVM), random forest(RF), XGBoost, and CatBoost, were analyzed. Responses were defined as ≥50 % depression improvement after treatment. Response rates between those with and without pretreatment prediction in another independent outpatient cohort (n = 208) were compared. Prediction accuracy using combined EEG features by SVM was better than frontal theta by logistic regression. The accuracy for OPD patients significantly dropped using the RCTD-trained SVM model. Modern ML models, especially RF (rTMS = 83.3 %, iTBS = 88.9 %, p -value(ACC > NIR) < 0.05 for iTBS), performed significantly above chance and had higher accuracy than SVM using both selected features (p < 0.05, FDR corrected for multiple comparisons) or all EEG features. Response rates among those receiving prediction before treatment were significantly higher than those without prediction (p = 0.035). The first study combining linear and non-linear EEG features could accurately predict responses to left PFC iTBS. The bootstraps-based ML model (i.e., RF) had the best predictive accuracy for rTMS and iTBS. • Frontal EEG activity may predict antidepressant responses to prefrontal rTMS or iTBS. • We compared prediction accuracy between different machine learning (ML) models and between different datasets. • Modern ML models, especially random forest, performed significantly above change. • Combing linear and non-linear EEG features accurately predict responses to rTMS and iTBS. [ABSTRACT FROM AUTHOR]

To increase leakage diagnosis (LD) efficiency of heating-networks and overcome the lack of actual leakage data, a LD method is proposed based on system simulation and principal component analysis (PCA)_back propagation (BP) neural network that treats the LD problem as a pattern recognition one. In this method, a hydraulic working-condition (HWC) simulation model is constructed to obtain model datasets (MD) of heating-network operation and HWC experiments are conducted to obtain experimental datasets (ED), then cross datasets (CD) is constructed by mixing MD and ED with different ratio. These three datasets are handled through the PCA to unify data feature distribution and realize feature transfer learning in neural network. Then two BP neural networks are trained by ED and CD respectively and both tested with ED. Finally, four kinds of experimental heating-networks, including a branch network with single heat source (B-SHS), a branch network with double heat sources (B-DHS), a single-ring network with single heat source (SR-SHS) and a double-ring network with double heat sources (DR-DHS), are researched. The results indicate that the LD prediction accuracy of BP neural network trained by CD is higher than that one trained by ED. The LD prediction accuracy varies with the cross-data ratio: from 97.52% to 92.21% for B-SHS, from 97.23% to 92.02% for B-DHS, from 97.85% to 89.74% for SR-SHS, and from 95.89% to 92.50% for DR-DHS. Furthermore, the proposed LD method based on BP neural network has better transfer learning performance than that based on random forest or support vector machine. [ABSTRACT FROM AUTHOR]

• Improve the classification accuracy for the area under crops. • Evaluation of the effect of data integration on classification accuracy. • Increase the computational efficiency of the models using SFS method. Timeliness, accurate, and frequent monitoring of area under crops are necessary for achieving food security. Therefore, the present study was conducted to improve the classification accuracy for the area under crops based on feature selection (FS) technique from multi-temporal Sentinel-1 and Sentinel-2 images in Shush County, Iran. To this end, the required pre-processing was performed on the images, and then features were extracted and categorized into four types of datasets with 78, 53, 65 features, for dataset 1 to 3 respectively. Dataset 4 is the result of integrating datasets 1, 2, and 3 with 194 features. The Sequential Forward Selection (SFS) method was used for the optimal FS, and Support Vector Machine (SVM) and Random Forest (RF) algorithms were used for the image classification. The results of FS by SFS method and SVM algorithm for datasets 1, 2, 3, and 4 showed that 15 features were optimal. The SFS method and RF algorithm determined the optimum feature number as 18, 19, 11, and 14 for datasets 1, 2, 3, and 4, respectively. The results of the RF algorithm performed on dataset 4 showed the highest Overall Accuracy (OA) and Kappa coefficient of 91.23% and 0.9, respectively. The results from this algorithm indicated that the type of input features affects the classification accuracy, so dataset 4 (compared to dataset 2) improved the classification with a difference in OA and kappa coefficient of approximately 2.5% and 0.03, respectively. According to the findings, it can be concluded that selecting the optimal number of features using the SFS method in the separation of area under crops is of high importance and improves the computational efficiency of the models and classification accuracy. The SFS result indicated that the higher accuracy could be achieved when less than 20 features recruited in classification. [ABSTRACT FROM AUTHOR]

To develop an intratumoural and peritumoural magnetic resonance imaging (MRI)-based radiomics nomogram for predicting tumour grade to improve clinical treatment and long-term prognosis. MRI (3 T) features and T2-weighted imaging with fat-saturation (T2WI-FS)-based radiomics features of 57 patients with soft-tissue sarcoma (STS) were analysed retrospectively. Tumour size, ratio of width and length, relative depth to the peripheral fascia, peritumoural oedema, heterogeneity on T2WI, necrosis signal, enhancement model, and peritumoural enhancement were obtained. Independent risk factors were screened to construct an MRI feature nomogram. Radiomics features were obtained from intratumoural and peritumoural images on T2WI-FS. The optimal radiomics model was selected by the four-step dimensionality reduction method of minimum and maximum normalisation, optimal feature selection, selection based on support vector machine with L1-norm regularisation model, and iterative feature selection. MRI features and optimal radiomics features were used to construct a radiomics nomogram. The MRI feature nomogram model, the radiomics model, and the radiomics nomogram model were assessed by receiver operating characteristic (ROC) curves and calibration curves of the training and validation sets. Heterogeneity on T2WI and peritumoural enhancement were independent risk factors for predicting high-grade STS. The areas under the curves of the training set and verification set of the three models were as follows: MRI feature nomogram, 0.86 and 0.83, respectively; intratumoural and peritumoural combined radiomics model, 0.99 and 0.86, respectively; and radiomics nomogram model, 0.98 and 0.96, respectively. The radiomics nomogram model based on MRI features and combined intratumoural and peritumoural radiomic features was best able to predict high-grade STS. • The relationship between peritumoral imaging and STS grading is reported first time. • The peripheral information of the tumor was obtained by expanding the radius by 5 mm. • The peritumor radiomic features are of great diagnostic value for high-grade STS. • The IPTV model is superior to the single model in predicting high grade STS. • IPTV radiomics nomogram perform well in predicting high-grade STS. [ABSTRACT FROM AUTHOR]

The goal of the work described here was to develop and assess a deep learning-based model that could automatically segment anterior chamber angle (ACA) tissues; classify iris curvature (I-Curv), iris root insertion (IRI), and angle closure (AC); automatically locate scleral spur; and measure ACA parameters in ultrasound biomicroscopy (UBM) images. A total of 11,006 UBM images were obtained from 1538 patients with primary angle-closure glaucoma who were admitted to the Eye Center of Renmin Hospital of Wuhan University (Wuhan, China) to develop an imaging database. The UNet++ network was used to segment ACA tissues automatically. In addition, two support vector machine (SVM) algorithms were developed to classify I-Curv and AC, and a logistic regression (LR) algorithm was developed to classify IRI. Meanwhile, an algorithm was developed to automatically locate the scleral spur and measure ACA parameters. An external data set of 1,658 images from Huangshi Aier Eye Hospital was used to evaluate the performance of the model under different conditions. An additional 439 images were collected to compare the performance of the model with experts. The model achieved accuracies of 95.2%, 88.9% and 85.6% in classification of AC, I-Curv and IRI, respectively. Compared with ophthalmologists, the model achieved an accuracy of 0.765 in classifying AC, I-Curv and IRI, indicating that its high accuracy was as high as that of the ophthalmologists (p > 0.05). The average relative errors (AREs) of ACA parameters were smaller than 15% in the internal data sets. Intraclass correlation coefficients (ICCs) of all the angle-related parameters were greater than 0.911. ICC values of all iris thickness parameters were greater than 0.884. The accurate measurement of ACA parameters partly depended on accurate localization of the scleral spur (p < 0.001). The model could effectively and accurately evaluate the ACA automatically based on fully automated analysis of UBM images, and it can potentially be a promising tool to assist ophthalmologists. The present study suggested that the deep learning model can be extensively applied to the evaluation of ACA and AC-related biometric risk factors, and it may broaden the application of UBM imaging in the clinical research of primary angle-closure glaucoma. [ABSTRACT FROM AUTHOR]

The effort to identify and comprehend potential earthquake-related phenomena shares a common goal: successful earthquake forecasting. Advancements in science and technology have made this goal multidisciplinary. Currently, possibly earthquake-related anomalies in the Vertical Total Electron Content (VTEC) of the Earth's ionosphere are being investigated. Global Navigation Satellite Systems (GNSS) can be used to calculate this ionospheric parameter. In this research work, GPS VTEC was calculated for periods between the years 2015 and 2019. The selection of this periods considered both seismically active and non-seismically areas in Mexico. The M w ≥ 5 earthquakes under study were registered by the National Seismological Service. Moreover, different geomagnetic storm and solar activity parameters, such as the geomagnetic equatorial Dst index and the F10.7 index, were analyzed. Additionally, the daily average and monthly mean number of sunspots (R, SSN, respectively) were included as a direct, long-term record of the development of the solar cycle. To the periods under study different statistical methods were applied, such as Mean-Square Error (MSE) and cross-correlation. The above aims to apply a machine learning technique capable of classifying between periods with seismic and non-seismic activity. The features were constructed using statistical data and results from the implemented analysis. Furthermore, Principal Component Analysis (PCA) was applied to reduce the feature vector dimensions, and accuracy scores were compute using k-fold cross-validation. The results from the Support Vector Machine (SVM) model indicated an accuracy of 88.9% for the training set, and an accuracy of 80% was obtained for the test set. One of the limitations of the current study was the sample size. However, the present initial approach for classifying seismic events from non-seismic periods using SVM demonstrated promising results when considering the indicated parameters and the days under study. [ABSTRACT FROM AUTHOR]

Treatment of locally advanced cervical cancer patients includes chemoradiation followed by brachytherapy. Our aim is to develop a delta radiomics (DRF) model from MRI-based brachytherapy treatment and assess its association with progression free survival (PFS). A retrospective analysis of FIGO stage IB- IV cervical cancer patients between 2012 and 2018 who were treated with definitive chemoradiation followed by MRI-based intracavitary brachytherapy was performed. Clinical factors together with 18 radiomic features extracted from different radiomics matrices were analyzed. The delta radiomic features (DRFs) were extracted from MRI on the first and last brachytherapy fractions. Support Vector Machine (SVM) models were fitted to combinations of 2–3 DRFs found significant after Spearman correlation and Wilcoxon rank sum test statistics. Additional models were tested that included clinical factors together with DRFs. A total of 39 patients were included in the analysis with a median patient age of 52 years. Progression occurred in 20% of patients (8/39). The significant DRFs using two DRF feature combinations was a model using auto correlation (AC) and sum variance (SV). The best performing three feature model combined mean, AC & SV. Additionally, the inclusion of FIGO stages with the 2- and 3 DRF combination model(s) improved performance compared to models with only DRFs. However, all the clinical factor + DRF models were not significantly different from one another (all AUCs were 0.77). Our study shows promising evidence that radiomics metrics are associated with progression free survival in cervical cancer. [ABSTRACT FROM AUTHOR]

This article presents a learning algorithm for dendrite morphological neurons (DMN) based on stochastic gradient descent (SGD). In particular, we focus on a DMN topology that comprises spherical dendrites, smooth maximum activation function nodes, and a softmax output layer, whose original learning algorithm is performed in two independent stages: (1) dendrites' centroids are learned by k-means, and (2) softmax layer weights are adjusted by gradient descent. A drawback of this learning method is that both stages are unplugged; once dendrites' centroids are defined, they keep static during weights learning, so no feedback is performed to correct the dendrites' positions to improve classification performance. To overcome this issue, we derive the delta rules for adjusting the dendrites' centroids and the output layer weights by minimizing the cross-entropy loss function under an SGD scheme. This gradient descent-based learning is feasible because the smooth maximum activation function that interfaces the dendrite units with the output layer is differentiable. The proposed DMN is compared against eight morphological neuron models with distinct topologies and learning methods and four well-established classifiers: support vector machine (SVM), multilayer perceptron (MLP), and random forest (RF), and k-nearest neighbors (k-NN). Besides, the classification performance is evaluated on 81 datasets. The experimental results show that the proposed method tends to outperform the DMN methods and is competitive or even better than SVM, MLP, RF, and k-NN. Thus, it is an alternative approach that can effectively be used for pattern classification. Moreover, SGD for DMN learning standardizes this neural model, like current artificial neural networks. [ABSTRACT FROM AUTHOR]

Twin support vector machine (TSVM) is a practical machine learning algorithm, whereas traditional TSVM can be limited for data with outliers or noises. To address this problem, we propose a novel TSVM with the symmetric LINEX loss function (SLTSVM) for robust classification. There are several advantages of our method: (1) The performance of the proposed SLTSVM for data with outliers or noise can be improved by using the symmetric LINEX loss function. (2) The introduction of regularization term can effectively improve the generalization ability of our model. (3) An efficient iterative algorithm is developed to solve the optimization problems of our SLTSVM. (4) The convergence and time complexity of the iterative algorithm are analyzed in detail. Furthermore, our model does not involve loss function parameter, which makes our method more competitive. Experimental results on synthetic, benchmark and image datasets with label noises and feature noises demonstrate that our proposed method slightly outperforms other state-of-the-art methods on most datasets. • This paper proposes SLTSVM by introducing symmetric LIENX loss function. • An interpretable method is proposed for selecting the loss parameter. • The convergence of the designed solution algorithm are further analyzed. • Experiments show that our method is efficient for data with outliers or noises. [ABSTRACT FROM AUTHOR]

A block valve installed along a trunk natural gas pipeline can automatically shut down when it detects a pressure-drop rate that is over a threshold value for a specified duration, indicating that leakage accidents have occurred. However, current methods have difficulties on identifying the differences among the pressure-drop rate signals caused by small-hole leakage, compressor startup, and block valve emergency shut-down conditions. This usually leads to improperly shutting down the block valve. Based on an optimized support vector machine, a machine-learning method is proposed to recognize the pressure-drop rate signals through the eigenvalue of the pressure-drop curve instead of only using the pressure-drop signal and its duration. Firstly, the singular decomposition method and the support vector machine method are applied to initially classify the pressure-drop rate signals of three operating conditions: the gas leakage, compressor startup, and block valve emergency shut-down conditions. Secondly, a teaching-learning-based optimization method is applied to optimize the penalty factor and kernel function parameter in the support vector machine model. Particularly, the tent chaotic maps and adaptive inertia weight methods are applied to improve the teaching-learning-based optimization method to balance its local and global search performance. Thirdly, an improved teaching-learning-based support vector machine method is established. Finally, 960 sets of simulated pressure-drop signals taken from three trunk natural gas pipelines and pressure-drop rate signals collected from an actual pipeline are applied to verify the accuracy of the proposed model. The results show that the improved teaching-learning-based support vector machine model achieved high classification and recognition accuracy. Specifically, it achieved accuracy rates of 99.4%, 100%, 98.3%, and 100% for the pressure-drop rate signals generated by pipelines A, B, C, and Cangzhou branch in the presence of pipeline leakage (the leak hole diameter is from 25 to 125 mm). Additionally, it demonstrated an average recognition accuracy of 97.42% for the pressure-drop rate signals generated by pipelines under other operating conditions. Through cooperating with Supervisory Control And Data Acquisition system, this method provides a more relevant approach to determine the shutdown conditions of a block valve while preventing mistaken actions. [ABSTRACT FROM AUTHOR]

With booming computer technology and diverse computer-based smart applications, intelligent monitoring and detection of the fatigue state of coal miners (miners) has attracted extensive attention from enterprises. In order to accurately and fast knowledge the fatigue status of Miners and reduce production accidents. The article proposes a Fatigue Monitoring and Detection (FMD) model based on a fusion Machine Learning (ML) approach: Marine Predators Algorithm (MPA)-optimized Least Squares Support Vector Machine (LSSVM). Firstly, the physiological information of electroencephalogram (ECG) of coal miners before and after manual handling operations was collected using the MP160 recorder (multiconductance physiological 160) produced by BIOPAC, USA. Using the paired-samples t-test method, the characteristic indicators reflecting miners' fatigue were extracted from the ECG. Secondly, Principal Component Analysis (PCA) was used to optimize the selected feature indicators and establish the depth fatigue feature parameter set to characterize the fatigue level of miners. Finally, the proposed MPA-LSSVM-based FMD model is applied to recognize Miners' fatigue levels. The results show that the selected indexes can effectively reflect the fatigue status of Miners. The proposed MPA-LSSVM-based FMD model has higher recognition accuracy than SVM and LSSVM models (13.99% and 18.68% higher, respectively) and better robustness. Therefore, the proposed MPA-LSSVM-based FMD model can accurately and effectively identify the fatigue status of Miners. [ABSTRACT FROM AUTHOR]

Obstructive sleep apnea (OSA) is a very prevalent disease and its diagnosis is based on polysomnography (PSG). We investigated whether snoring-sound-, very low frequency electrocardiogram (ECG-VLF)- and thoraco-abdominal effort- PSG signal entropy values could be used as surrogate markers for detection of OSA and OSA severity classification. The raw data of the snoring-, ECG- and abdominal and thoracic excursion signal recordings of two consecutive full-night PSGs of 86 consecutive patients (22 female, 53.74 ± 12.4 years) were analyzed retrospectively. Four epochs (30 s each, manually scored according to the American Academy of Sleep Medicine standard) of each sleep stage (N1, N2, N3, REM, awake) were used as the ground truth. Sampling entropy (SampEn) of all the above signals was calculated and group comparisons between the OSA severity groups were performed. In total, (86x4x5 =)1720 epochs/group/night were included in the training set as an input for a support vector machine (SVM) algorithm to classify the OSA severity classes. Analyses were performed for first- and second-night PSG recordings separately. Twenty-seven patients had mild (RDI = ≥ 5/h but <15/h), 21 patients moderate (RDI ≥15/h but <30/h) and 23 patients severe OSA (RDI ≥30/h). Fifteen patients had an RDI <5/h and were therefore considered non-OSA. Using SE on the above three PSG signal data and using a SVM pipeline, it was possible to distinguish between the four OSA severity classes. The best metric was snoring signal-SE. The area-under-the-curve (AUC) calculations showed reproducible significant results for both nights of PSG. The second night data were even more significant, with non-OSA (R) vs. light OSA (L) 0.61, R vs. moderate (M) 0.68, R vs. heavy OSA (H) 0.84, L vs. M 0.63, M vs. H 0.65 and L vs. H 0.82. The results were not confounded by age or gender. SampEn of either snoring-, very low ECG-frequencies- or thoraco-abdominal effort signals alone may be used as a surrogate marker to diagnose OSA and even predict OSA severity. More specifically, in this exploratory study snoring signal SampEn showed the greatest predictive accuracy for OSA among the three signals. Second night data showed even more accurate results for all three parameters than first-night recordings. Therefore, technologies using only parts of the PSG signal, e.g. sound-recording devices, may be used for OSA screening and OSA severity group classification. • The entropy of various PSG biosignals can be used to facilitate OSA diagnosis. • The Sampling entropy of the snoring signal on polysomnography accurately helps to diagnose and classify OSA. • Entropy analysis could replace the complex and costly PSG in the future. [ABSTRACT FROM AUTHOR]

• A framework of fault prognosis strategy of external gear pump with insufficient availability of experimental data. • Generation of the high-fidelity CFD model to tackle the absence/ minimal availability of experimental data. • Generation of various working conditions of the pump to comprehend its behaviour. • Implementation of synthetic data generation methods to increase data availability and to generate degradation behaviour. • Noise sensitivity analysis of ML algorithms by adding white noise in the training and test dataset. Fault prognosis is an important area of research that aims to predict and diagnose faults in complex systems. The sudden failure of industrial components can have adverse consequences for an organisation in terms of time, cost and workflow. It is, therefore, critical to ensure the maintenance of equipment components in optimal condition in order to avoid downtime that may cause significant disruption. Due to this reason, in recent years, there has been an increasing interest in creating innovative methods for fault prognosis that can increase system reliability and reduce maintenance costs. Gear pumps are widely utilised in a variety of industrial applications, and their reliability and effectiveness are crucial for achieving optimal system performance. Gear pumps, on the other hand, are prone to malfunctions and failures, which can result in substantial downtime and maintenance costs. The challenge is to develop a fault prognosis approach that is reliable and accurate enough to detect and diagnose defects in a gear pump before they cause system failures. This paper proposes a novel computational strategy for the fault prognosis of an external gear pump using Machine Learning (ML) approaches. Due to the unavailability of sufficient experimental data in the vicinity of failure mechanisms, a novel approach to generating a high-fidelity in-silico dataset via a Computational Fluid Dynamic (CFD) model of the gear pump in healthy and faulty working conditions is presented. However, considering the computational demand for rerunning the same CFD simulations, novel synthetic data generation techniques are implemented by perturbing the frequency content of the time series to recreate other working conditions and constructing degradation behaviour using linear and cubic interpolation methods for run-to-failure scenarios. The synthetically created datasets are used to train the underlying ML metamodel for fault prognosis. Two types of ML algorithms are employed for fault prognosis: Multilayer Perceptron (MLP) and Support Vector Machine (SVM) algorithms. A series of numerical examples are shown, allowing us to infer that the proposed modelling technique is feasible in an industrial setting and that employing the MLP algorithm delivers superior fault prognosis results when compared to SVM. Furthermore, datasets generated using the cubic interpolation method have lower prediction errors than datasets generated using the linear interpolation method due to the smoother degradation behaviour in the data. [ABSTRACT FROM AUTHOR]

Metal organic frameworks (MOFs) have been studied vastly for hydrogen storage purposes due to their unique properties. In the present work hydrogen storage capacity is modeled using machine learning approaches including Multi-layer Perceptron (MLP), Support Vector Machines (SVM), Random Forest (RF), CatBoost, LightGBM, XGBoost, and Committee Machine Intelligence System (CMIS) trained on experimental data gathered from various experimental studies. CMIS model shows the best results with R 2 = 0.982 and Root Mean Square Error = 0.088. Additionally, this study uses the Leverage method to detect and eliminate suspected data points in order to improve predictability. Only 2.04% of data points are detected as suspected data points. Sensitivity analysis demonstrates that surface area, pore volume, pressure, and temperature have almost the same contribution on hydrogen storage. The findings of this research offer a valuable model for MOF selection and optimization of operational conditions in hydrogen storage processes that utilizes MOFs. [Display omitted] • Machine learning models are developed to predict the H 2 storage capacity of MOFs. • CMIS shows the best results with R 2 = 0.982 and RMSE = 0.088. • The Leverage approach is used to recognize and eliminate suspected data points. • Sensitivity analysis showed equal contribution of model's inputs. [ABSTRACT FROM AUTHOR]

• Accuracy of avalanche hazard analysis and prediction rely on avalanche feedback. • Convention methods of avalanche reporting are inefficient and incomplete. • UAV RGB images can provide accurate avalanche occurrence details of an area. • Detected deposits will help in avalanche analysis, characterization and modelling. Avalanche activities in the snow-covered mountainous slopes are very frequent and stochastically distributed over time and geographical extents. Conventional methods of avalanche occurrence reporting in these areas are inefficient and incomplete due to inhospitable weather and inaccessible topographic conditions. This paper presents a novel framework for the detection and analysis of avalanche deposits using UAV (Unmanned Aerial Vehicle) RGB images of snow-bound avalanche-prone areas. In the first part of the proposed framework, an OBIA-CNN (Object Based Image Analysis-Convolution Neural Network) method is applied to UAV images for the detection of avalanche deposits. Initially, the OBIA-based multi-scale multi-resolution segmentation technique is used for UAV image segmentation, then the classification of these image segments was performed using a CNN classifier to detect the avalanche deposits. The CNN-based classifier was created, trained, validated, and applied to generate the class probability maps of different classes present in the images. The post-classification refinements were then applied to the detected deposits. In the second step the detailed accuracy analysis of the detected deposits based on their count, area, size similarity, and shape similarity with respect to reference deposits has been carried out to establish the suitability of the proposed method. The avalanche deposits detected using the proposed method were found in good correlation with manually delineated deposits in terms of all four accuracy criteria. The precision of the model was found 1.0, the value of recall was 0.88 and the F-1 score was 0.93. The overall object shape matching accuracy of the detected deposits was also found high (0.92). The results of the proposed deposit detection method were also compared with the classification results of Support Vector Machine (SVM), Random Forest (RF), and UNET to evaluate the performance of the method. Finally, the surface area and snow volume-based characterization of the detected deposits was performed. The proposed framework will be useful for the automated detection of avalanche deposits and their characterization for the regions of specific interest. The study suggests that regular monitoring of avalanche activities using UAV images provides a good solution for the operational monitoring of important corridors of snowbound regions. [ABSTRACT FROM AUTHOR]

Monitoring valve operation status is very significant in saving natural gas resources and realizing sustainability of the fossil energy. At present, many machine learning algorithms have been applied to nondestructive testing (NDT) field. Among them, the twin support vector machine (TWSVM) is a representative binary classification method. However, the traditional TWSVM assigns same classification weights to all sample points, including classification boundary points and other non-boundary sample points. These same classification weights often lead to classification errors or overfitting in TWSVM. Therefore, to overcome the drawback, we propose an improved nonlinear TWSVM (I-TWSVM) to optimize the classification weights assignment processing. In new assignment processing, the boundary points are given greater classification weights than other non-boundary points, which makes the I-TWSVM becomes more sensitive than the traditional TWSVM to boundary points. The I-TWSVM has been applied to recognize the acoustic emission (AE) sample data for valve internal leak fault accident. The leak recognition experiment revealed that the proposed algorithm is better than the traditional nonlinear TWSVM in classification accuracy and sensitivity, and its calculation time is faster than that of nonlinear TWSVM, which can promote the nondestructive testing technology development for low-pressure natural gas valve. [ABSTRACT FROM AUTHOR]

• It is possible to obtain good classification results of cardiovascular autonomic neuropathy using supervised machine learning. • The best classifier is the random forest classifier with a F1-score of 60% on the reduced dataset. • In the ensembled data no standard clinical biochemical analytes were present in the 10 best classifiers. Using supervised machine learning to classify the severity of cardiovascular autonomic neuropathy (CAN). The aims were 1) to investigate which features contribute to characterising CAN 2) to generate an ensembled set of features that best describes the variation in CAN classification. Eighty-two features from demographic, beat-to-beat, biochemical, and inflammation were obtained from 204 people with diabetes and used in three machine-learning-classifiers, these are: support vector machine, decision tree, and random forest. All data were ensembled using a weighted mean of the features from each classifier. The 10 most important features derived from the domains: Beat-to-beat, inflammation markers, disease-duration, and age. Beat-to-beat measures associate with CAN as diagnosis is mainly based on cardiac reflex responses, disease-duration and age are also related to CAN development throughout disease progression. The inflammation markers may reflect the underlying disease process, and therefore, new treatment modalities targeting systemic low-grade inflammation should potentially be tested to prevent the development of CAN. Cardiac reflex responses should be monitored closely to diagnose and classify severity levels of CAN accurately. Standard clinical biochemical analytes, such as glycaemic level, lipidic level, or kidney function were not included in the ten most important features. Beat-to-beat measures accounted for approximately 60% of the features in the ensembled data. [ABSTRACT FROM AUTHOR]

Electroencephalography (EEG) is a supplementary diagnostic tool in psychiatry but lacks practical usage. EEG has demonstrated inconsistent diagnostic ability because major depressive disorder (MDD) is a heterogeneous psychiatric disorder with complex pathologies. In clinical psychiatry, it is essential to detect these complexities using multiple EEG paradigms. Though the application of machine learning to EEG signals in psychiatry has increased, an improvement in its classification performance is still required clinically. We tested the classification performance of multiple EEG paradigms in drug-naïve patients with MDD and healthy controls (HCs). Thirty-one drug-naïve patients with MDD and 31 HCs were recruited in this study. Resting-state EEG (REEG), the loudness dependence of auditory evoked potentials (LDAEP), and P300 were recorded for all participants. Linear discriminant analysis (LDA) and support vector machine (SVM) classifiers with t -test-based feature selection were used to classify patients and HCs. The highest accuracy was 94.52 % when 14 selected features, including 12 P300 amplitudes (P300A) and two LDAEP features, were layered. The accuracy was 90.32 % when a SVM classifier for 30 selected features (14 P300A, 14 LDAEP, and 2 REEG) was layered in comparison to each REEG, P300A, and LDAEP, the best accuracies of which were 71.57 % (2-layered with LDA), 87.12 % (1-layered with LDA), and 83.87 % (6-layered with SVM), respectively. The present study was limited by small sample size and difference in formal education year. Multiple EEG paradigms are more beneficial than a single EEG paradigm for classifying drug-naïve patients with MDD and HCs. • Patients with depression showed higher P300s and lower LDAEPs than healthy individuals. • Patients with depression showed decreased resting state EEG absolute power in high alpha and delta bands. • Classification performance using multiple EEG paradigms was 94.52 % between patients and healthy individuals. [ABSTRACT FROM AUTHOR]

In the realm of multi-class classification, the twin K-class support vector classification (Twin-KSVC) generates ternary outputs { − 1 , 0 , + 1 } by evaluating all training data in a "1-versus-1-versus-rest" structure. Recently, inspired by the least-squares version of Twin-KSVC and Twin-KSVC, a new multi-class classifier called improvements on least-squares twin multi-class classification support vector machine (ILSTKSVC) has been proposed. In this method, the concept of structural risk minimization is achieved by incorporating a regularization term in addition to the minimization of empirical risk. Twin-KSVC and its improvements have an influence on classification accuracy. Another aspect influencing classification accuracy is feature selection, which is a critical stage in machine learning, especially when working with high-dimensional datasets. However, most prior studies have not addressed this crucial aspect. In this study, motivated by ILSTKSVC and the cardinality-constrained optimization problem, we propose ℓ p -norm least-squares twin multi-class support vector machine (PLSTKSVC) with 0 < p < 1 to perform classification and feature selection at the same time. The technique employed to solve the optimization problems associated with PLSTKSVC is user-friendly, as it involves solving systems of linear equations to obtain an approximate solution for the proposed model. Under certain assumptions, we investigate the properties of the optimum solutions to the related optimization problems. Several real-world datasets were tested using the suggested method. According to the results of our experiments, the proposed method outperforms all current strategies in most datasets in terms of classification accuracy while also reducing the number of features. [ABSTRACT FROM AUTHOR]

Support vector machines (SVMs) are powerful statistical learning tools, but their application to large datasets can cause time-consuming training complexity. To address this issue, various instance selection (IS) approaches have been proposed, which choose a small fraction of critical instances and screen out others before training. However, existing methods have not been able to balance accuracy and efficiency well. Some methods miss critical instances, while others use complicated selection schemes that require even more execution time than training with all original instances, thus violating the initial intention of IS. In this work, we present a newly developed IS method called Valid Border Recognition (VBR). VBR selects the closest heterogeneous neighbors as valid border instances and incorporates this process into the creation of a reduced Gaussian kernel matrix, thus minimizing the execution time. To improve reliability, we propose a strengthened version of VBR (SVBR). Based on VBR, SVBR gradually adds farther heterogeneous neighbors as complements until the Lagrange multipliers of already selected instances become stable. In numerical experiments, the effectiveness of our proposed methods is verified on benchmark and synthetic datasets in terms of accuracy, execution time and inference time. [ABSTRACT FROM AUTHOR]

Multi-view learning aims to make use of the advantages of different views to complement each other and fully mines the potential information in the data. However, the complexity of multi-view learning algorithm is much higher than that of single view learning algorithm. Based on the optimality conditions of two classical multi-view models: SVM-2K and multi-view twin support vector machine (MvTwSVM), this paper analyzes the corresponding relationship between dual variables and samples, and derives their safe screening rules for the first time, termed as SSR-SVM-2K and SSR-MvTwSVM. It can assign or delete four groups of different dual variables in advance before solving the optimization problem, so as to greatly reduce the scale of the optimization problem and improve the solution speed. More importantly, the safe screening criterion is "safe", that is, the solution of the reduced optimization problem is the same as that of the original problem before screening. In addition, we further give a sequence screening rule to speed up the parameter optimization process, and analyze its properties, including the similarities and differences of safe screening rules between multi-view SVMs and single-view SVMs, the computational complexity, and the relationship between the parameter interval and screening rate. Numerical experiments verify the effectiveness of the proposed methods. [ABSTRACT FROM AUTHOR]

Machine learning has been shown to outperform traditional statistical methods for risk prediction model development. We aimed to develop machine learning-based risk prediction models for cardiovascular mortality and hospitalisation for ischemic heart disease (IHD) using self-reported questionnaire data. The 45 and Up Study was a retrospective population-based study in New South Wales, Australia (2005–2009). Self-reported healthcare survey data on 187,268 participants without a history of cardiovascular disease was linked to hospitalisation and mortality data. We compared different machine learning algorithms, including traditional classification methods (support vector machine (SVM), neural network, random forest and logistic regression) and survival methods (fast survival SVM, Cox regression and random survival forest). A total of 3687 participants experienced cardiovascular mortality and 12,841 participants had IHD-related hospitalisation over a median follow-up of 10.4 years and 11.6 years respectively. The best model for cardiovascular mortality was a Cox survival regression with L1 penalty at a re-sampled case/non-case ratio of 0.3 achieved by under-sampling of the non-cases. This model had the Uno's and Harrel's concordance indexes of 0.898 and 0.900 respectively. The best model for IHD hospitalisation was a Cox survival regression with L1 penalty at a re-sampled case/non-case ratio of 1.0 with Uno's and Harrel's concordance indexes of 0.711 and 0.718 respectively. Machine learning-based risk prediction models developed using self-reported questionnaire data had good prediction performance. These models may have the potential to be used in initial screening tests to identify high-risk individuals before undergoing costly investigation. • Risk prediction models were developed for CVD mortality and IHD hospitalisation. • These models were developed using machine learning methods. • These models used exclusively self-reported variables for risk prediction. • These models may allow for cost-effective CVD risk screening and monitoring. [ABSTRACT FROM AUTHOR]

• There was a decreasing trend in absolute power spectral density of EEG changes induced by auditory stimuli with decreasing levels of consciousness. • This cortical response was effective in distinguishing between good and bad prognoses for patients with consciousness disorders. • Cortical responses to auditory stimulation may be a significant electrophysiological indicator of prognosis in patients with disorders of consciousness. This study aimed to assess the prognosis of patients with disorders of consciousness (DoC) using auditory stimulation with electroencephalogram (EEG) recordings. We enrolled 72 patients with DoC in the study, which involved subjecting patients to auditory stimulation while EEG responses were recorded. Coma Recovery Scale–Revised (CRS-R) scores and Glasgow Outcome Scale (GOS) were determined for each patient and followed up for three months. A frequency spectrum analysis was performed on the EEG recordings. Finally, the power spectral density (PSD) index was used to predict the prognosis of patients with DoC based on a support vector machine (SVM) model. Power spectral analyses revealed that the cortical response to auditory stimulation showed a decreasing trend with decreasing consciousness levels. Auditory stimulation-induced changes in absolute PSD at the delta and theta bands were positively correlated with the CRS-R and GOS scores. Furthermore, these cortical responses to auditory stimulation had a good ability to discriminate between good and poor prognoses of patients with DoC. Auditory stimulation-induced changes in the PSD were highly predictive of DoC outcomes. Our findings showed that cortical responses to auditory stimulation may be an important electrophysiological indicator of prognosis in patients with DoC. [ABSTRACT FROM AUTHOR]

In recent years, functional analysis of brain networks based on graph theory properties has attracted considerable attention. This approach has usually been exploited for structural and functional brain analysis, while its potential in motor decoding tasks has remained unexplored. This study aimed to investigate the feasibility of using graph-based features in hand direction decoding in movement execution and preparation intervals. Hence, EEG signals were recorded from nine healthy subjects while performing a four-target center-out reaching task. The functional brain network was calculated based on the magnitude-squared coherence (MSC) at six frequency bands. Then, the features based on eight graph theory metrics were extracted from brain networks. The classification was performed with a support vector machine classifier. The results revealed that in four-class direction discrimination, the mean accuracy of the graph-based method surpassed 63% and 53% on movement and pre-movement data, respectively. Additionally, a feature fusion approach that combines the graph theory features with power features was proposed. The fusion method raised the classification accuracy to 70.8% and 61.2% for movement and pre-movement intervals, respectively. This work has verified the feasibility of using graph theory properties and their superiority over band power features in a hand movement decoding task. • Investigating the feasibility of using graph theory analysis in movement decoding. • Both movement and pre-movement intervals were investigated. • Graph theory metrics attain superior results compare to band power features. • The fusion approach of graph metrics with power features led to best results. [ABSTRACT FROM AUTHOR]

• An improved singular spectrum decomposition is proposed. • High complexity components are secondarily decomposed to reduce its complexity. • An improved prediction model of optimization algorithm is proposed. • Intelligent weighting strategy is introduced to overcome traditional weighting problem. • A combined prediction model for carbon emission is proposed. Accurate prediction of carbon emission is critical for the development of low-carbon economy. However, most carbon emission prediction studies use a single model with low prediction accuracy, and do not consider the instability of carbon emission. Therefore, this paper proposes a combined prediction model of carbon emission. Firstly, the original data is decomposed by singular spectrum decomposition to obtain a limited amount of singular spectrum components. Secondly, high complexity components are secondarily decomposed by variational mode decomposition. Then, chameleon swarm algorithm and carnivorous plant algorithm are used to train the regularization coefficients and kernel parameters of kernel extreme learning machine and least squares support vector machine respectively, and the trained model is used to predict the decomposition components. Finally, induced ordered weighted averaging operator is used to calculate the weight of single model, and error correction is introduced to further promote the prediction accuracy. The carbon emission data of China and the United States is used to make a prediction experiment. The results indicate that the proposed model is superior to other comparative models in different indexes, which provides a new idea for carbon emission prediction. [ABSTRACT FROM AUTHOR]