Search Results (1,482)

Background: Decoding motor intentions from electroencephalogram (EEG) signals is a critical component of motor imagery-based brain–computer interface (MI–BCIs). In traditional EEG signal classification, effectively utilizing the valuable information contained within the electroencephalogram is crucial. Objectives: To further optimize the use of information from various domains, we propose a novel framework based on multi-domain feature rotation transformation and stacking ensemble for classifying MI tasks. Methods: Initially, we extract the features of Time Domain, Frequency domain, Time-Frequency domain, and Spatial Domain from the EEG signals, and perform feature selection for each domain to identify significant features that possess strong discriminative capacity. Subsequently, local rotation transformations are applied to the significant feature set to generate a rotated feature set, enhancing the representational capacity of the features. Next, the rotated features were fused with the original significant features from each domain to obtain composite features for each domain. Finally, we employ a stacking ensemble approach, where the prediction results of base classifiers corresponding to different domain features and the set of significant features undergo linear discriminant analysis for dimensionality reduction, yielding discriminative feature integration as input for the meta-classifier for classification. Results: The proposed method achieves average classification accuracies of 92.92%, 89.13%, and 86.26% on the BCI Competition III Dataset IVa, BCI Competition IV Dataset I, and BCI Competition IV Dataset 2a, respectively. Conclusions: Experimental results show that the method proposed in this paper outperforms several existing MI classification methods, such as the Common Time-Frequency-Spatial Patterns and the Selective Extract of the Multi-View Time-Frequency Decomposed Spatial, in terms of classification accuracy and robustness. Full article

The current process of embryo selection in in vitro fertilization is based on morphological criteria; embryos are manually evaluated by embryologists under subjective assessment. In this study, a deep learning-based pipeline was developed to classify the viability of embryos using combined inputs, including microscopic images of embryos and additional features, such as patient age and developed pseudo-features, including a continuous interpretation of Istanbul grading scores by predicting the embryo stage, inner cell mass, and trophectoderm. For viability prediction, convolution-based transferred learning models were employed, multiple pretrained models were compared, and image preprocessing techniques and hyperparameter optimization via Optuna were utilized. In addition, a custom weight was trained using a self-supervised learning framework known as the Simple Framework for Contrastive Learning of Visual Representations (SimCLR) in cooperation with generated images using generative adversarial networks (GANs). The best model was developed from the EfficientNet-B0 model using preprocessed images combined with pseudo-features generated using separate EfficientNet-B0 models, and optimized by Optuna to tune the hyperparameters of the models. The designed model’s F1 score, accuracy, sensitivity, and area under curve (AUC) were 65.02%, 69.04%, 56.76%, and 66.98%, respectively. This study also showed an advantage in accuracy and a similar AUC when compared with the recent ensemble method. Full article

(1) Background: The exploration of various machine learning (ML) algorithms for classifying the state of Lumbar Intervertebral Discs (IVD) in orthopedic patients is the focus of this study. The classification is based on six key biomechanical features of the pelvis and lumbar spine. Although previous research has demonstrated the effectiveness of ML models in diagnosing IVD pathology using imaging modalities, there is a scarcity of studies using biomechanical features. (2) Methods: The study utilizes a dataset that encompasses two classification tasks. The first task classifies patients into Normal and Abnormal based on their IVDs (2C). The second task further classifies patients into three groups: Normal, Disc Hernia, and Spondylolisthesis (3C). The performance of various ML models, including decision trees, support vector machines, and neural networks, is evaluated using metrics such as accuracy, AUC, recall, precision, F1, Kappa, and MCC. These models are trained on two open-source datasets, using the PyCaret library in Python. (3) Results: The findings suggest that an ensemble of Random Forest and Logistic Regression models performs best for the 2C classification, while the Extra Trees classifier performs best for the 3C classification. The models demonstrate an accuracy of up to 90.83% and a precision of up to 91.86%, highlighting the effectiveness of ML models in diagnosing IVD pathology. The analysis of the weight of different biomechanical features in the decision-making processes of the models provides insights into the biomechanical changes involved in the pathogenesis of Lumbar IVD abnormalities. (4) Conclusions: This research contributes to the ongoing efforts to leverage data-driven ML models in improving patient outcomes in orthopedic care. The effectiveness of the models for both diagnosis and furthering understanding of Lumbar IVD herniations and spondylolisthesis is outlined. The limitations of AI use in clinical settings are discussed, and areas for future improvement to create more accurate and informative models are suggested. Full article

The issue of power quality disturbances in modern power systems has become increasingly complex and severe, with multiple disturbances occurring simultaneously, leading to a decrease in the recognition accuracy of traditional algorithms. This paper proposes a composite power quality disturbance identification method based on the integration of improved Complementary Ensemble Empirical Mode Decomposition (CEEMDAN), Hilbert Transform (HT), and Extreme Learning Machine (ELM). Addressing the limitations of traditional signal processing techniques in handling nonlinear and non-stationary signals, this study first preprocesses the collected initial power quality signals using the improved CEEMDAN method to reduce modal aliasing and spurious components, thereby enabling a more precise decomposition of noisy signals into multiple Intrinsic Mode Functions (IMFs). Subsequently, the HT is utilized to conduct a thorough analysis of the reconstructed signals, extracting their time-amplitude information and instantaneous frequency characteristics. This feature information provides a rich data foundation for subsequent classification and identification. On this basis, an improved ELM is introduced as the classifier, leveraging its powerful nonlinear mapping capabilities and fast learning speed to perform pattern recognition on the extracted features, achieving accurate identification of composite power quality disturbances. To validate the effectiveness and practicality of the proposed method, a simulation experiment is designed. Upon examination, the approach introduced in this study retains a fault diagnosis accuracy exceeding 95%, even amidst significant noise disturbances. In contrast to conventional techniques, such as Convolutional Neural Network (CNN) and Support Vector Machine (SVM), this method achieves an accuracy enhancement of up to 5%. Following optimization via the Particle Swarm Optimization (PSO) algorithm, the model’s accuracy is boosted by 3.6%, showcasing its favorable adaptability. Full article

Mangrove ecosystems are increasingly threatened by climate change and coastal development, making precise ecological stress modeling essential for informing conservation strategies. This study employs AI-based classification techniques to classify mangroves using Landsat 8-SR OLI/TIRS sensors (2023) along the Jazan Coast, identifying a total mangrove area of 19.4 km². The ensemble classifier achieved an F1 score of 95%, an overall accuracy of 93%, and a kappa coefficient of 0.86. Ecological stress was modeled via a generalized additive model (GAM) with key predictors, including trends in the NDVI, NDWIveg (vegetation water content), NDWIow (open water), and LST from 1991 to 2023, which were derived using surface reflectance (SR) products from Landsat 5 TM, Landsat 7 ETM+, and Landsat 8 OLI/TIRS sensors. The model exhibited strong performance, with an R² of 0.89. Model diagnostics using linear regression (R² = 0.86), a high F-statistic, minimal intercept, and 10-fold cross-validation confirmed the model’s robustness, with a consistent MSE (0.12) and cross-validated R² of 0.86. Moran’s I analysis also indicated significant spatial clustering. Findings indicate that mangroves in non-ravine, mainland coastal areas experience more ecological stress from disruptions in freshwater and sediment supply due to recent developments. In contrast, island coastal areas exhibit low stress levels due to minimal human activity, except in dense canopy regions where significant stress, likely linked to climate change, was observed. These results underscore the need for further investigation into the drivers of this ecological pressure. Full article

In this paper, a novel set of 13 handcrafted features derived from the contours of ships in synthetic aperture radar (SAR) images is introduced for ship classification. Additionally, the information entropy is presented as a valuable metric for quantifying the confidence (or uncertainty) associated with classification predictions. Two segmentation methods for the contour extraction were investigated: a classical approach using the watershed algorithm and a U-Net architecture. The features were tested using a support vector machine (SVM) on the OpenSARShip and FUSAR-Ship datasets, demonstrating improved results compared to existing handcrafted features in the literature. Alongside the SVM, a random forest (RF) and a Gaussian process classifier (GPC) were used to examine the effect of entropy derivation from different classifiers while assessing feature robustness. The results show that when aggregating predictions of an ensemble, techniques such as entropy-weighted averaging are shown to produce higher accuracies than methods like majority voting. It is also found that the aggregation of individual entropies within an ensemble leads to a normal distribution, effectively minimizing outliers. This characteristic was utilized to model the entropy distributions, from which confidence levels were established based on Gaussian parameters. Predictions were then assigned to one of three confidence levels (high, moderate, or low), with the Gaussian-based approach showing superior correlation with classification accuracy compared to other methods. Full article

Fault detection and classification in transmission lines are critical for maintaining the reliability and stability of electrical power systems. Quick and accurate fault detection allows for timely intervention, minimizing equipment damage, and reducing downtime. This study addresses the challenge of effective fault classification, particularly when dealing with smaller, more practical datasets. Initially, the study examined the performance of conventional machine learning algorithms on a comprehensive dataset of 7681 samples, demonstrating high accuracy owing to the inherent symmetry of sinusoidal voltage and current signals. However, the true efficacy of these algorithms was evaluated by minimizing the dataset to 231 training samples, with the remainder being used for testing. A novel Multi-Target Ensemble Classifier was developed to improve classification accuracy. The proposed algorithm achieved an impressive overall accuracy of 0.829165, outperforming traditional methods, including the K-Nearest Neighbors Classifier, support vector classification, random forest classifier, decision tree classifier, AdaBoost classifier, gradient boosting classifier, and Gaussian NB. This research highlights the importance of efficient fault classification techniques in power systems and proposes a superior solution in the form of a multitarget ensemble classifier. Full article

Ensemble classification is widely used in multi-label algorithms, and it can be divided into homogeneous ensembles and heterogeneous ensembles according to classifier types. A heterogeneous ensemble can generate classifiers with better diversity than a homogeneous ensemble and improve the performance of classification results. An Adaptive Heterogeneous Classifier Group (AHCG) algorithm is proposed. The AHCG first proposes the concept of a Heterogeneous Classifier Group (HCG); that is, two groups of different ensemble classifiers are used in the testing and training phases. Secondly, the Adaptive Selection Strategy (ASS) is proposed, which can select the ensemble classifiers to be used in the test phase. The least squares method is used to calculate the weights of the base classifiers for the in-group classifiers and dynamically update the base classifiers according to the weights. A large number of experiments on seven datasets show that this algorithm has better performance than most existing ensemble classification algorithms in terms of its accuracy, example-based F1 value, micro-averaged F1 value, and macro-averaged F1 value. Full article

Background/Objectives: Noninvasive brain stimulation (NIBS) can boost motor recovery after a stroke. Certain movement phases are more responsive to NIBS, so a system that auto-detects these phases would optimize stimulation timing. This study assessed the effectiveness of various machine learning models in identifying movement phases in hemiparetic individuals undergoing simultaneous NIBS and EEG recordings. We hypothesized that transcranial direct current stimulation (tDCS), a form of NIBS, would enhance EEG signals related to movement phases and improve classification accuracy compared to sham stimulation. Methods: EEG data from 10 chronic stroke patients and 11 healthy controls were recorded before, during, and after tDCS. Eight machine learning algorithms and five ensemble methods were used to classify two movement phases (hold posture and reaching) during each of these periods. Data preprocessing included z-score normalization and frequency band power binning. Results: In chronic stroke participants who received active tDCS, the classification accuracy for hold vs. reach phases increased from pre-stimulation to the late intra-stimulation period (72.2% to 75.2%, p < 0.0001). Late active tDCS surpassed late sham tDCS classification (75.2% vs. 71.5%, p < 0.0001). Linear discriminant analysis was the most accurate (74.6%) algorithm with the shortest training time (0.9 s). Among ensemble methods, low gamma frequency (30–50 Hz) achieved the highest accuracy (74.5%), although this result did not achieve statistical significance for actively stimulated chronic stroke participants. Conclusions: Machine learning algorithms showed enhanced movement phase classification during active tDCS in chronic stroke participants. These results suggest their feasibility for real-time movement detection in neurorehabilitation, including brain–computer interfaces for stroke recovery. Full article

Background: Psoriasis is a chronic, immune-mediated skin disease characterized by lifelong persistence and fluctuating symptoms. The clinical similarities among its subtypes and the diversity of symptoms present challenges in diagnosis. Early diagnosis plays a vital role in preventing the spread of lesions and improving patients’ quality of life. Methods: This study proposes a hybrid model combining multiple transfer learning and ensemble learning methods to classify psoriasis subtypes accurately and efficiently. The dataset includes 930 images labeled by expert dermatologists from the Dermatology Clinic of Fırat University Hospital, representing four distinct subtypes: generalized, guttate, plaque, and pustular. Class imbalance was addressed by applying synthetic data augmentation techniques, particularly for the rare subtype. To reduce the influence of nonlesion environmental factors, the images underwent systematic cropping and preprocessing steps, such as Gaussian blur, thresholding, morphological operations, and contour detection. DenseNet-121, EfficientNet-B0, and ResNet-50 transfer learning models were utilized to extract feature vectors, which were then combined to form a unified feature set representing the strengths of each model. The feature set was divided into 80% training and 20% testing subsets and evaluated using a hard voting classifier consisting of logistic regression, random forest, support vector classifier, k-nearest neighbors, and gradient boosting algorithms. Results: The proposed hybrid approach achieved 93.14% accuracy, 96.75% precision, and an F1 score of 91.44%, demonstrating superior performance compared to individual transfer learning models. Conclusions: This method offers significant potential to enhance the classification of psoriasis subtypes in clinical and real-world settings. Full article

This paper introduces a novel approach for classifying multidimensional physiological and clinical data using Synolitic Graph Neural Networks (SGNNs). SGNNs are particularly good for addressing the challenges posed by high-dimensional datasets, particularly in healthcare, where traditional machine learning and Artificial Intelligence methods often struggle to find global optima due to the “curse of dimensionality”. To apply Geometric Deep Learning we propose a synolitic or ensemble graph representation of the data, a universal method that transforms any multidimensional dataset into a network, utilising only class labels from training data. The paper demonstrates the effectiveness of this approach through two classification tasks: synthetic and fMRI data from cognitive tasks. Convolutional Graph Neural Network architecture is then applied, and the results are compared with established machine learning algorithms. The findings highlight the robustness and interpretability of SGNNs in solving complex, high-dimensional classification problems. Full article

This study addresses the critical challenge of the limited understanding of environmental factors influencing cocoa cultivation in Colombia, a region with significant production potential but diverse agroecological conditions. The fragmented nature of the existing agricultural data and the lack of targeted research hinder efforts to optimize productivity and sustainability. To bridge this gap, this research employs a data-driven approach, using advanced machine learning techniques such as supervised, unsupervised, and ensemble models, to analyze environmental datasets and provide actionable recommendations. By integrating data from official Colombian sources, as well as the NASA POWER database, and geographical APIs, the present study proposes a methodology to systematically assess environmental conditions and classify regions for optimal cocoa cultivation. The use of an assembled model, combining clustering with targeted machine learning for each cluster, offers a more precise and scalable understanding of cocoa establishment under diverse conditions. Despite challenges such as limited dataset resolution and localized climate variability, this research provides valuable insights for a more comprehensive understanding of the environmental conditions impacting cocoa plantation establishment in a given location. The key findings reveal that temperature, humidity, and wind speed are crucial determinants of cocoa growth, with complex interactions affecting regional suitability. The results offer valuable guidance for the implementation of adaptive agricultural practices and resilience strategies, enabling sustainable cocoa production systems. By implementing better practices, countries such as Colombia can achieve higher market shares under growing global cocoa demand conditions. Full article

Rice is a staple food for almost half of the world’s population, and the stability and sustainability of rice production plays a decisive role in food security. Diseases are a major cause of loss in rice crops. The timely discovery and control of diseases are important in reducing the use of pesticides, protecting the agricultural eco-environment, and improving the yield and quality of rice crops. Deep convolutional neural networks (DCNNs) have achieved great success in disease image classification. However, most models have complex network structures that frequently cause problems, such as redundant network parameters, low training efficiency, and high computational costs. To address this issue and improve the accuracy of rice disease classification, a lightweight deep convolutional neural network (DCNN) ensemble method for rice disease classification is proposed. First, a new lightweight DCNN model (called CG-EfficientNet), which is based on an attention mechanism and EfficientNet, was designed as the base learner. Second, CG-EfficientNet models with different optimization algorithms and network parameters were trained on rice disease datasets to generate seven different CG-EfficientNets, and a resampling strategy was used to enhance the diversity of the individual models. Then, the sequential least squares programming algorithm was used to calculate the weight of each base model. Finally, logistic regression was used as the meta-classifier for stacking. To verify the effectiveness, classification experiments were performed on five classes of rice tissue images: rice bacterial blight, rice kernel smut, rice false smut, rice brown spot, and healthy leaves. The accuracy of the proposed method was 96.10%, which is higher than the results of the classic CNN models VGG16, InceptionV3, ResNet101, and DenseNet201 and four integration methods. The experimental results show that the proposed method is not only capable of accurately identifying rice diseases but is also computationally efficient. Full article

In the current work, authors reviewed the latest research results in landslide susceptibility mapping (LSM) using artificial intelligence (AI) methods. Based on an overall review of collected publications, the review was classified into four sections based on their complexity: single-model approaches, enhanced models with optimization, ensemble models, and hybrid models. Each category offers distinct advantages and is suited to specific geographic and data conditions, enabling the selection of an optimal model type based on the complexity and requirements of the mapping task. Among models, random forest (RF), support vector machine (SVM), convolutional neural network (CNN), and multilayer perception (MLP) are used as the baseline to compare any new model introduced to develop LSM. Moreover, compared to previous review works, the number of LSM conditioning factors used in AI models are significantly increased, up to 122 factors. Their relation to the AI models is illustrated using Sankey diagram, while a radar chart is used to further visualize the dataset size per reviewed work for comparative purposes. In the main part of the current review work, the main findings are summarized into a table form, where the reader can find the overall relations between landslide conditioning factors, landslide dataset size, applied AI models, and their accuracy on predicting LSM for selected geographical locations. In terms of the regions, Asia is leading in the application of AI models to generate LSM, and in such regions with dense populations falling into higher landslide risk categories, there are more ongoing research activities, using modern AI methods. This trend underscores the increased use of AI in disaster management, with implications for improving practical applications, such as early warning systems and informing policy decisions aimed at risk reduction in vulnerable areas. Full article

This study developed an advanced ensemble learning model aimed to improve the accuracy of predicting sarcopenia, a condition characterized by a gradual decline in muscle mass and strength, leading to increased disability and mortality. The study focused on enhancing model performance by combining various machine learning methods and addressing critical challenges, such as class imbalance and data complexity. Several foundational models were employed, including support vector machine, random forest, neural network, logistic regression, and decision tree. To address class imbalance, the adaptive synthetic sampling method was implemented, producing synthetic samples for the minority class to achieve a more balanced dataset. The data preprocessing stage included feature scaling and feature selection processes, utilizing recursive feature elimination to refine feature selection. Subsequently, a classifier selection algorithm was employed to select models that provided an optimal balance of diversity and performance. The effectiveness of the final ensemble model was evaluated using various metrics, such as accuracy, precision, recall, F1-score, and ROC AUC. The model achieved an accuracy of 88.5%, outperforming individual machine learning models and existing methods in the literature. These findings suggest that the classifier selection algorithm effectively addresses challenges in sarcopenia prediction, particularly in the case of imbalanced data. The model’s strong performance indicates its potential for use in clinical environments, where it can facilitate early diagnosis and improve intervention strategies for sarcopenia patients. This study advances the field of medical machine learning by demonstrating the utility of ensemble learning in healthcare prediction. Full article