This study aims to develop a hybrid system for the automatic annotation of scientific texts that efficiently processes multilingual publications using state-of-the-art natural language processing (NLP) technologies. The system integrates classical algorithms (Gensim, NLTK) with transformer-based models via the Cohere API to achieve high semantic consistency and accuracy in annotations. The system architecture comprises modules for data acquisition, preprocessing, manual and automatic annotation, data storage, and quality control. The performance of the proposed model was benchmarked against established methods such as BERTSUM, TF-IDF + LSA, and GPT-3.5-turbo using evaluation metrics including ROUGE, BLEU, and METEOR. The hybrid model outperformed other automated systems, demonstrating superior scores across ROUGE-1 (0.52), BLEU (0.41), and METEOR (0.39) metrics, indicating its effectiveness in producing concise and semantically accurate summaries. The system also achieved 100% language detection accuracy and 90% accuracy in semantic word relationships via Word2Vec. The integration of traditional statistical methods with advanced transformer models enables the proposed system to deliver high-quality annotations suitable for diverse scientific domains. The results validate the model’s ability to process and summarize complex scientific texts effectively. This system provides a scalable, secure, and user-friendly platform for researchers, institutions, and developers. It supports multilingual annotation, seamless API integration, and potential deployment in cloud environments, offering significant benefits for academic, biomedical, and information-intensive sectors.

Risk stratification of impaired glycemic control remains a major challenge in biomedical data analysis due to heterogeneous metabolic, behavioral, and therapeutic factors observed in large-scale populations. This study proposes a calibrated and interpretable decision–support framework, termed Calibrated Multi-Task Stacking Ensemble (CMSE), for joint modeling of clinically related glycemic outcomes. The framework integrates demographic variables, lipid profiles, renal and inflammatory biomarkers, dietary and smoking indicators, and therapy-related features within a unified predictive architecture. Robust modeling is ensured through leakage-aware preprocessing, quantile-based Winsorization, out-of-fold stacking, and isotonic calibration of probabilistic outputs. The physiological coherence between short-term and long-term glycemic markers is investigated using an explicit intertask coupling mechanism based on the estimated average glucose (eAG) ratio. Model interpretability is supported using SHAP analysis, mutual information, distance correlation, and feature importance metrics. In the primary medication-free screening configuration, the framework is evaluated on the NHANES 2017–March 2020 dataset, achieving ROC-AUC of 0.865 for diabetes classification and R² values of 0.385 and 0.366 for plasma glucose and HbA1c prediction, respectively. These results indicate that CMSE provides a reliable and explainable approach for calibrated glycemic risk assessment and clinical decision support.

Early detection of autism spectrum disorder (ASD) is essential for enabling timely interventions that significantly impact developmental outcomes. Traditional diagnostic procedures often rely on clinical observations and subjective assessments, which can reduce objectivity and delay diagnosis. This study presents an automated classification approach for detecting ASD in children using gait kinematic data obtained through video-based skeletal tracking. We propose a hybrid deep learning architecture that combines one-dimensional convolutional layers (Conv1D), bidirectional long short-term memory (BiLSTM) networks, and a multi-head attention mechanism to capture complex spatiotemporal patterns in motion data. The dataset includes 100 participants—50 diagnosed with ASD and 50 typically developing (TD) peers. Preprocessing steps included Euclidean norm transformation, logarithmic scaling, Z-score normalization, and sliding window segmentation. The proposed model achieved 94.9% accuracy, 0.91 recall, 0.86 precision, and an area under the curve (AUC) of 0.97, outperforming a baseline long short-term memory (LSTM) architecture. These findings demonstrate the potential of gait-based kinematic features and hybrid neural networks for objective and reproducible ASD screening. The developed system can contribute to AI-assisted decision support tools in clinical and educational environments, enhancing diagnostic accuracy, and supporting inclusive developmental care.

Objective and reproducible diagnosis of attention-deficit/hyperactivity disorder (ADHD) remains challenging because of the limited availability of reliable electroencephalography (EEG) biomarkers and the high variability of neural signals. This study proposes a computational framework for ADHD detection based on dynamic functional connectivity and network-index modeling. Multichannel EEG recordings were transformed into temporal connectivity graphs using sliding-window correlations of band-limited amplitude envelopes. Several network-index models were evaluated, including linear, graph-based, recurrent, and hybrid spatio-temporal approaches. The proposed Hybrid Spatio-Temporal Graph Transformer demonstrated moderate, yet reproducible, subject-level classification performance. On the independent test set, the model achieved an accuracy of 63.16%, a balanced accuracy of 62.22%, a sensitivity of 80.00%, a specificity of 44.44%, an F1-score of 69.57%, and an AUC-ROC of 0.7444. Additional analysis of the derived network index demonstrated moderate intergroup separability, with a mean index shift of 1.16, Cohen’s d = 0.73, Pearson’s r = 0.36, and distribution overlap = 0.72. These findings suggest that the proposed framework captures informative spatio-temporal EEG connectivity patterns associated with ADHD; however, the model’s diagnostic applicability should be considered preliminary and requires validation in larger independent cohorts.

The early and objective detection of neurodevelopmental conditions in children remains a critical challenge due to the reliance on subjective behavioral assessments and clinician-dependent interpretation. This study proposes a hybrid deep learning framework for the automated analysis of electroencephalographic (EEG) signals aimed at identifying neurodevelopmental patterns associated with attention and cognitive regulation disorders. The methodology integrates convolutional neural networks (CNN), gated recurrent units (GRU), and long short-term memory (LSTM) layers to extract both spatial and temporal EEG features. Preprocessing procedures include band-pass filtering, independent component analysis, segmentation into epochs, and z-score normalization. Experimental evaluation demonstrates high classification accuracy and robust generalization ability. Additionally, explainable artificial intelligence techniques are employed to interpret feature importance at the electrode level, increasing clinical transparency and trust. The obtained results confirm that hybrid deep learning architectures combined with interpretability tools can serve as reliable decision-support mechanisms in pediatric neurodiagnostics.

This paper discusses the application of machine learning methods for the automatic detection of erosion zones based on the NDVI index. The U-Net model with the EfficientNetB0 pre-trained encoder is used, which allows us to achieve high segmentation accuracy. The study includes the preparation and analysis of geospatial data, model training, and testing on data from the region of South Kazakhstan. The developed system demonstrates an accuracy of 99.99%, which confirms the effectiveness of the proposed methodology. The obtained results can be used for monitoring soil degradation and taking measures to prevent erosion processes.