Mach. Learn. Knowl. Extr., Volume 8, Issue 3 (March 2026) – 30 articles

Visual localization plays a critical role for mobile robots to estimate their position and orientation in GPS-denied environments. However, its efficiency, robustness, and generalization are fundamentally undermined by severe viewpoint changes and dramatic appearance variations, which present persistent challenges for image-based feature representation and pose estimation under real-world conditions. Recently, map-free visual relocalization (MFVR) has emerged as a promising paradigm for lightweight deployment and privacy isolation on edge devices, while how to learn compact and invariant image tokens without relying on structural 3D maps still remains a core problem, particularly in highly dynamic or long-term scenarios. In this paper, we propose the Debiased Multiplex Tokenizer as a novel method (termed as DMT-Loc) for efficient and versatile MFVR to address these issues. Specifically, DMT-Loc is built upon a pretrained vision Mamba encoder and integrates three key modules for relative pose regression: First, Multiplex Interactive Tokenization yields robust image tokens with non-local affinities and cross-domain descriptions. Second, Debiased Anchor Registration facilitates anchor token matching through proximity graph retrieval and autoregressive pointer attribution. Third, Geometry-Informed Pose Regression empowers multi-layer perceptrons with a symmetric swap gating mechanism operating inside each decoupled regression head to support accurate and flexible pose prediction in both pair-wise and multi-view modes. Extensive evaluations across seven public datasets demonstrate that DMT-Loc substantially outperforms existing baselines and ablation variants in diverse indoor and outdoor environments. Full article

Background: Real-time and accurate analysis of breast ultrasounds is crucial for diagnosis but remains challenging due to issues like low image contrast and operator dependency. This study aims to address these challenges by developing an integrated framework for real-time lesion detection and tracking. Methods: The proposed system combines Contrast-Limited Adaptive Histogram Equalization (CLAHE) for image preprocessing, a transfer learning-enhanced YOLOv11 model following a continual learning paradigm for cross-center generalization in for lesion detection, and a novel Detection-Based Tracking (DBT) approach that integrates Kernelized Correlation Filters (KCF) with periodic detection verification. The framework was evaluated on a dataset comprising 11,383 static images and 40 ultrasound video sequences, with a subset verified through biopsy and the remainder annotated by two radiologists based on radiological reports. Results: The proposed framework demonstrated high performance across all components. The transfer learning strategy (TL12) significantly improved detection outcomes, achieving a mean Average Precision (mAP) of 0.955, a sensitivity of 0.938, and an F1 score of 0.956. The DBT method (KCF + YOLO) achieved high tracking accuracy, with a success rate of 0.984, an Intersection over Union (IoU) of 0.85, and real-time operation at 54 frames per second (FPS) with a latency of 7.74 ms. The use of CLAHE preprocessing was shown to be a critical factor in improving both detection and tracking stability across diverse imaging conditions. Conclusions: This research presents a robust, fully integrated framework that bridges the gap between speed and accuracy in breast ultrasound analysis. The system’s high performance and real-time efficiency underscore its strong potential for clinical adoption to enhance diagnostic workflows, reduce operator variability, and improve breast cancer assessment. Full article

Applications for machine learning (ML) and deep learning (DL) are constantly growing and have already been adopted in the field of particle measurement technology. Even though analytical (ultra-)centrifugation (AC/AUC) is a widely used technique for characterizing dispersed particle systems, ML and DL have not yet been applied in this area. Data evaluation and interpretation in AC/AUC can be challenging and often requires expert knowledge. DL models can help, but their development is limited by a lack of annotated training data. One solution is to generate and use synthetic data instead. In the first part of this study, a model was trained to synthesize data from experiments using a combination of Variational Autoencoder (VAE) and Generative Adversarial Networks (GANs). The results appear highly realistic. Novice users could distinguish real from synthetic samples with only 63% accuracy. Then, a classifier was trained on experimental AC data to categorize real-world examples based on their underlying separation kinetics, testing different DL architectures. After initial training, the models were further fine-tuned with synthetic AC data. ResNet34 models achieved the best performance with 94% accuracy, comparable to an AC expert (91%), while inexperienced users reached only 53%. In the second part of our study, a regression model was trained for the analysis of sedimentation coefficients. Therefore, various generative models were developed and evaluated for synthesizing AUC data based on numerically simulated sedimentation boundaries. The best results were achieved by combining VAE and GAN architectures with embedded physical constraints. However, the generative networks have so far led to additional smearing of the profiles, resulting in a broadening of the sedimentation coefficient distribution and indicating that further refinement is necessary. Full article

The development of AI-driven computational methods has enabled rapid and non-invasive analysis of respiratory sounds using acoustic data, particularly cough recordings. Although the COVID-19 pandemic accelerated research on cough-based acoustic analysis, many early studies were limited by insufficient data quality, lack of standardized protocols, and limited reproducibility due to data scarcity. In this study, we propose an audio analysis framework for cough-based respiratory disease screening research using COVID-19 as a clinically validated case dataset. All analyses were conducted on a single clinically acquired multicentric dataset collected under standardized conditions in certified laboratories in Mexico and Spain, comprising cough recordings from 1105 individuals. Model training and testing were performed exclusively within this dataset. The framework incorporates signal preprocessing and a comparative evaluation of segmentation strategies, showing that segmented cough analysis significantly outperforms full-signal analysis. Class imbalance was addressed using the Synthetic Minority Over-sampling Technique (SMOTE) for CNN2D models and the supervised Resample filter implemented in WEKA for classical machine learning models, both applied exclusively to the training subset to generate balanced training sets and prevent data leakage. Feature extraction and classification were carried out using Random Forest, Support Vector Machine (SVM), XGBoost, and a 2D Convolutional Neural Network (CNN2D), with hyperparameter optimization via AutoML. The proposed framework achieved a best balanced screening performance of 85.58% sensitivity and 86.65% specificity (Random Forest with GeMAPSvB01), while the highest-specificity configuration reached 93.90% specificity with 18.14% sensitivity (CNN2D with SMOTE and AutoML). These results demonstrate the methodological feasibility of the proposed framework under the evaluated conditions. Full article

Emotion classification is a fundamental task in affective computing, with applications in human–computer interaction, mental health monitoring, and social media analysis. Although most existing methods formulate it as a flat classification problem, emotional expressions are inherently structured and grounded in semantic roles such as the emotion cue, stimulus, experiencer, and target. However, the relative contribution of these roles to emotion inference has not been systematically examined. Unlike prior models, we propose RolEmo, a role-aware framework for emotion classification that explicitly incorporates semantic role information. The framework employs a controlled role-masking strategy to analyze the contribution of individual roles, augments textual representations with external commonsense knowledge to capture implicit affective context, and applies supervised contrastive learning to structure the embedding space by bringing emotionally similar instances closer while separating opposing ones. We evaluate RolEmo on three benchmark datasets annotated with semantic roles. Experimental results demonstrate that RolEmo outperforms the strongest baseline across three datasets by up to $16.4\%$, $25.8\%$, and $23.2\%$ in the Full Text, Only Role, and Without Role settings, respectively. The analysis further indicates that the cue and stimulus roles provide the most reliable signals for emotion classification, with their removal causing performance drops of up to $6.2\%$ in macro f1-score, while experiencer and target roles exhibit more variable effects. These findings highlight the importance of structured semantic modeling and commonsense reasoning for robust and interpretable emotion understanding. Full article

In complex and unknown processes, global models are fitted over the entire input domain but often tend to perform poorly whenever the response surface exhibits non-stationary behavior and varying smoothness. A common approach is to use local models, which requires partitioning the input domain into subdomains and training multiple models, thereby adding significant complexity. Recognizing this limitation, this study addresses the need for models that represent the input–output relationship consistently over the full domain while still adapting to local variations in the response. It introduces a novel machine learning approach: the Polynomial Chaos Expanded Gaussian Process (PCEGP), leveraging polynomial chaos expansion to calculate input-dependent hyperparameters of the Gaussian process (GP). This provides a mathematically interpretable approach that incorporates non-stationary covariance functions and heteroscedastic noise estimation to generate locally adapted models. The model performance is compared to different algorithms in benchmark tests for regression tasks. The results demonstrate low prediction errors of the PCEGP, highlighting model performance that is often competitive with or better than previous methods. A key advantage of the presented model is its interpretable hyperparameters along with training and prediction runtimes comparable to those of a standard GP. Full article

Manual single-slice lumbar quantitative computed tomography (QCT) depends on operator-driven slice selection and trabecular region-of-interest (ROI) placement. We developed a fully automated single-slice workflow for vertebral trabecular Hounsfield unit (HU) measurement that combines unsuitable-slice prescreening, dual-purpose segmentation, intra-patient slice-quality ranking, and a deterministic inner ROI rule. The pipeline includes an Eligibility Gate, QC-Envelope segmentation for broad, vertebral- and usability-preserving delineation, PairRank-Swin for best-slice selection, and dedicated trabecular segmentation for final quantitative analysis. In the independent external cohort, 4 cases were considered non-evaluable by both manual review and the pipeline, and 2 additional borderline-quality cases were manually measured but rejected by the pipeline; therefore, paired HU agreement analysis included 44 evaluable cases. Agreement remained high, with Pearson’s r = 0.987, Lin’s CCC = 0.985, mean bias −0.44 HU, and limits of agreement from −14.88 to +13.99 HU. Coverage was 84.1% within ±10 HU and 97.7% within ±15 HU. Ablation analysis showed that slice ranking and ROI erosion were the most critical components. In an open module-level baseline comparison, QC-Envelope segmentation substantially outperformed TotalSegmentator. This workflow provides high agreement with expert HU measurement while preserving reviewable intermediate outputs. Full article

Infant cries serve as a primary indicator of distress and pain; however, distinguishing pain-related cries from those triggered by other needs remains a challenging task, even for trained professionals. Timely and accurate pain assessment is essential for appropriate medical intervention, particularly in preverbal infants who cannot express their needs verbally. Recently, Deep Learning (DL) models have demonstrated significant potential in addressing this challenge by enabling automated and efficient pain assessment through audio signal processing. In this survey, we review methods for pain assessment from infant cry sounds, covering deep learning architectures, modern Transformer-based models, and emerging Vision-Language Model (VLM) pipelines. The review includes approaches that integrate Mel-spectrogram representations of cry audio with multimodal model frameworks to improve robustness, interpretability, and cross-modal reasoning in pain detection. By summarizing recent advancements and identifying limitations and open challenges in current methodologies, this review aims to provide insights into future research directions that may enhance the robustness, generalizability, and clinical applicability of automated infant pain assessment tools. Full article

We propose a centralized Activity Retrieval and Detection Intelligence (ActivityRDI) solution framework, demonstrate its application performance in network threat detection in detail, and show its generalization to other domains. Network threat detection is challenging owing to the complex nature of attack activities and the limited historically revealed threat data from which to learn. To help enhance the existing methods (e.g., analytics, machine learning, and artificial intelligence) to detect the network threats, we propose a multi-agent AI solution for agile threat detection. In this solution, a knowledge graph is used to analyze changes in user activity patterns and calculate the risk of unknown threats. Then, an imbalanced learning model is used to prune and weight the knowledge graph and to calculate the risk of known threats. Finally, a large language model (LLM) is used to retrieve and interpret the risk associated with user activities from the knowledge graph and the imbalanced learning model. The preliminary results show that the solution improves the threat capture rate by 3–4% and adds natural language interpretations of the risk predictions based on user activities with 95% accuracy. Furthermore, a demonstration application has been built to show how the proposed solution framework can be deployed and used. The generalizability of the proposed solution in other domains is also shown through an application to customer engagement, with 97% accuracy. Full article

Large language models (LLMs) show promise for grading open-ended assessments but still exhibit inconsistent accuracy, systematic biases, and limited reliability across assignments. To address these concerns, we introduce SURE (Selective Uncertainty-based Re-Evaluation), a human-in-the-loop pipeline that combines repeated LLM prompting, uncertainty-based flagging, and selective human regrading. Three LLMs—gpt-4.1-nano, gpt-5-nano, and the open-source gpt-oss-20b—graded answers of 46 students to 130 open questions and coding exercises across five assignments. Each student answer was scored 20 times to derive majority-voted predictions and self-consistency-based certainty estimates. We simulated human regrading by flagging low-certainty cases and replacing them with scores from four human graders. We used the first assignment as a training set for tuning certainty thresholds and to explore LLM output diversification via sampling parameters, rubric shuffling, varied personas, multilingual prompts, and post hoc ensembles. We then evaluated the effectiveness and efficiency of SURE on the other four assignments using a fixed certainty threshold. Across assignments, fully automated grading with a single prompt resulted in substantial underscoring, and majority-voting based on 20 prompts improved but did not eliminate this bias. Low certainty (i.e., high output diversity) was diagnostic of incorrect LLM scores, enabling targeted human regrading that improved grading accuracy while reducing manual grading time by 40–90%. Aggregating responses from all three LLMs in an ensemble improved certainty-based flagging and most consistently approached human-level accuracy, with 70–90% of the grades students would receive falling inside human-grader ranges. A reanalysis based on outputs from a more diversified LLM ensemble comprised of gpt-5, codestral-25.01, and llama-3.3-70b-instruct replicated these findings but also suggested that large reasoning models such as gpt-5 might eliminate the need for human oversight of LLM grading entirely. These findings demonstrate that self-consistency-based uncertainty estimation and selective human oversight can substantially improve the reliability and efficiency of AI-assisted grading. Full article

Recent work on the evaluation of large language models emphasizes that the relevant unit of intelligence is not the artificial system alone but the human–AI hybrid. In parallel, topological and dynamical models of cognition based on Painlevé equations and non-semisimple topology propose that consciousness, intelligence, and creativity emerge from constrained long-horizon dynamics near criticality. This perspective article argues that these two research directions are deeply compatible. We show that the empirical framework for human–AI collaboration can be interpreted as a fusion process between complementary cognitive sectors: exploration (AI) and selection (human cognition). The dynamical mechanism underlying this fusion is identified with noisy phase locking between cognitive oscillators. Two independent routes to a universal $1/f$ spectral signature are developed: a geometric route through the WKB/Stokes analysis of Painlevé V confluence, and an arithmetic route through the Mangoldt function and harmonic interactions in phase-locked loops. We connect these results to the Bost–Connes quantum statistical model, whose phase transition at the pole of the Riemann zeta function provides an exact mathematical framework for the lock-in phase hypothesis of identity consolidation in AI systems. This synthesis suggests a unified research program for hybrid intelligence grounded in topology, dynamical systems, number theory, and real-world AI evaluation. Full article

Active matter actively searches for models that allow them to connect the behavior of multiple agents to particle system with a physical law. However, the arsenal of models used to model active matter systems is very restricted. Modern differential equation discovery approaches allow one to extract governing equations from data for a single particle in the form of the ODE. However, there is still the question of how to model at the meso- and macroscales. This paper presents a data-driven framework for extracting the governing physical laws of a hardware-made swarm across multiple scales of organization. Using the EPDE framework, we transition from a discrete, chaotic trajectory of individual agents to a continuous, effective field theory of the collective. We show that augmenting the symbolic search space with interaction-aware tokens allowed for the derivation of stochastic partial differential equations (SDEs) that significantly outperformed baseline deterministic models (reducing CRPS by up to 10%). Additionally, we derive a system of SPDEs that governs the macroscale displacement field. Full article

This study presents a novel data-driven framework for developing airport-specific landing policies and procedures from historical successful-landing data. The proposed process, termed the Airport-Dependent Landing Procedure (ADLP), is motivated by the fact that airports rely on uniquely tailored approach charts reflecting local operational constraints and environmental conditions. While existing approach charts and landing procedures are primarily designed based on expert knowledge, safety margins, and regulatory conventions, the authors argue that data science and data mining techniques offer a complementary and empirically grounded methodology for extracting operationally meaningful structures directly from historical landing data. In this work, we construct a probabilistic three-dimensional environment from real-world aircraft approach trajectories, capturing spatiotemporal relationships under varying atmospheric conditions during approach. The proposed methodology integrates Adversarial Inverse Reinforcement Learning (AIRL) with Recurrent Proximal Policy Optimization (R-PPO) to establish a foundation for automated landing without pilot intervention. AIRL infers reward functions that are consistent with behaviors exhibited in prior successful landings. Subsequently, R-PPO is employed to learn control policies that satisfy safety constraints related to airspeed, sink rate, and runway alignment. Application of the proposed framework to real approach trajectories at Guam International Airport demonstrates the efficiency and effectiveness of the methodology. Full article

The increasing complexity and heterogeneity of Internet of Things (IoT) systems pose significant challenges for systematic security and vulnerability assessment. From a knowledge-centric perspective, IoT security analysis requires transforming heterogeneous asset information into structured and interpretable security knowledge. In this paper, we propose a structured methodology for vulnerability analysis that models the attack surface of an IoT system by explicitly linking asset characteristics to known vulnerabilities, security controls, and countermeasures. The approach starts with a visual representation of the system architecture, where hardware, software, and communication components are identified and described through their technical characteristics. These characteristics are automatically mapped to relevant vulnerabilities, security controls, and countermeasures using a dedicated software tool called AVCA (Asset Vulnerabilities and Countermeasures Analyzer). The tool generates graph-based analytical representations that model vulnerabilities–countermeasures relationships in compliance with the Cloud Security Alliance (CSA) IoT Security Framework. From these graphs, attack–countermeasure trees are derived to provide a clear and interpretable representation of potential threats and mitigation strategies. The proposed methodology was evaluated through a case study involving a representative IoT system and an exploratory applicability experiment with participants with different levels of experience in IoT and cybersecurity. The results suggest that the approach is feasible and practically applicable for supporting security analysts in the systematic assessment of IoT attack surfaces, vulnerability identification, and selection of appropriate countermeasures under the evaluated conditions. This work highlights the role of structured and interpretable knowledge extraction as a foundation for knowledge-centric and interpretable IoT security analysis. Full article

Human–AI collaboration (HAIC) increasingly mediates high-risk decisions in public and private sectors, yet many documented AI harms arise not only from model error but from breakdowns in joint human–AI work: miscalibrated reliance, impaired contestability, misallocated agency, and governance opacity. Conventional explainable AI (XAI) approaches, often delivered as static one-shot artifacts, are poorly matched to these sociotechnical dynamics. This paper is a position paper arguing that explainability should be reframed as a harm-mitigation infrastructure for HAIC: an interactive, iterative capability that supports ongoing sensemaking, safe handoffs of control, governance stakeholder roles and institutional accountability. We introduce co-explainers as a conceptual framework for interactive XAI, in which explanations are co-produced through structured dialogue, feedback, and governance-aware escalation (explain → feedback → update → govern). To ground this position, we synthesize prior harm taxonomies into six HAIC-oriented harm clusters and use them as heuristic design lenses to derive cluster-specific explainability requirements, including uncertainty communication, provenance and logging, contrastive “why/why-not” and counterfactual querying, role-sensitive justification, and recourse-oriented interaction protocols. We emphasize that co-explainers do not “mitigate” sociotechnical harms in isolation; rather, they provide an interface layer that makes harms more detectable, decisions more contestable, and accountability handoffs more operational under realistic constraints such as sealed models, dynamic updates, and value pluralism. We conclude with an agenda for evaluating co-explainers and aligning interactive XAI with governance frameworks in real-world HAIC deployments. Full article

Dynamic characterization of calcareous (carbonate) sands is essential for performance-based design of offshore foundations, coastal reclamation, and marine infrastructure in tropical and subtropical regions. In contrast to silica sands, carbonate sediments are biogenic and typically comprise angular, irregular grains with intra-particle voids and fragile skeletal microstructure. These traits promote grain crushing and fabric evolution at relatively low-to-moderate confinement, leading to pronounced stress dependency, strong nonlinearity with strain amplitude, and substantial scatter in laboratory stiffness and damping measurements. Consequently, empirical correlations calibrated primarily on quartz sands may yield biased estimates when transferred to carbonate environments. This study presents an ML-driven, leakage-aware benchmarking framework for predicting two key dynamic parameters of biogenic calcareous sands, damping ratio $D$ and shear modulus $G$, using standard tabular descriptors commonly available in geotechnical practice. Two consolidated experimental databases were curated from resonant column and cyclic triaxial measurements ($D$: $n=890$; $G$: $n=966$), spanning mean effective confining stress $25 \le {\sigma }_m^{\prime }\le 1600 \mathrm{k}$Pa and a wide range of density and gradation conditions. To emphasize transferability, explicit deposit/site labels were excluded, and missingness arising from heterogeneous reporting was handled through a consistent preprocessing pipeline (training-only imputation, categorical encoding, and scaling). Eleven regression algorithms were evaluated, covering linear baselines, regularized regression, neighborhood learning, single trees, bagging and boosting ensembles, kernel regression, and a feedforward neural network. Performance was assessed using $R^2$, RMSE, and MAE on training/validation/test splits, and engineering credibility was supported through explainability-based diagnostics to verify mechanically plausible sensitivities. Results show that ensemble-tree models (Extra Trees and Random Forest) provide the most reliable accuracy–robustness balance across both targets, consistently outperforming linear models and the tested SVR configuration and exhibiting stable validation-to-test behavior. The explainability audit confirms physically meaningful separation of governing controls: stiffness is primarily stress-controlled (${\sigma }_m^{\prime }$ dominant for $G$), whereas damping is primarily strain-controlled ($\gamma$ dominant for $D$). The proposed framework supports practical deployment as a fast surrogate for generating $G\hspace{0.17em}-\hspace{0.17em}\gamma$ and $D\hspace{0.17em}-\hspace{0.17em}\gamma$ curves within the training domain and for guiding targeted laboratory test planning in carbonate settings. Full article

The rapid growth of Internet of Medical Things (IoMT) devices has revolutionized diagnostics and patient care within smart healthcare networks. However, this progress has also expanded the attack surface due to the heterogeneity and interconnectivity of medical devices. To overcome the limitations of traditional batch-trained security models, this study proposes an adaptive online intrusion detection framework designed for real-time operation in dynamic healthcare environments. The system combines Leveraging Bagging with Hoeffding Tree classifiers for incremental learning while integrating the Page–Hinkley test to detect and adapt to concept drift in evolving attack patterns. A modular and scalable network architecture supports centralized monitoring and ensures seamless interoperability across various IoMT protocols. Implemented within a low-latency, high-throughput stream-processing pipeline, the framework meets the stringent clinical requirements for responsiveness and reliability. To simulate streaming conditions, we evaluated the model using the CICIoMT2024 dataset, presenting one instance at a time in random order to reflect dynamic, real-time traffic in IoMT networks. Experimental results demonstrate exceptional performance, achieving accuracies of 0.9963 for binary classification, 0.9949 for six-class detection, and 0.9860 for nineteen-class categorization. These results underscore the framework’s practical efficacy in protecting modern healthcare infrastructures from evolving cyber threats. Full article

We forecast monthly Canadian real GDP growth using machine learning models trained on Official macroeconomic indicators and Google Trends (GT) data. Predictors are selected dynamically in each rolling window using PDC-SIS, with cross-validation-based tuning to support real-time forecasting and avoid data leakage. The evaluation is conducted on the latest-available (final-vintage) series and should be interpreted as a pseudo out-of-sample forecasting exercise rather than real-time vintage nowcasting. We evaluate GBM, XGBoost, LightGBM, CatBoost, and Random Forest against an ARIMA baseline. Official data deliver the strongest performance at short and medium horizons, while combining Official and GT data yields the clearest improvement at the longest horizon. With GT data alone, LightGBM is the only ML model maintaining positive out-of-sample R² across all horizons. Diebold–Mariano tests corroborate these patterns: LightGBM dominates other ML models under GT-only predictors, whereas with Official and combined data, the horizon-specific best models significantly outperform ARIMA, with smaller differences among leading tree-based methods. Full article

Selecting a source language for zero-shot cross-lingual transfer is typically done by intuition or by defaulting to English, despite large performance differences across language pairs. We study whether linguistic similarity can predict transfer performance and support principled source-language selection. We introduce quantified WALS (qWALS), a typology-based similarity metric derived from features in the World Atlas of Language Structures, and evaluate it against existing similarity baselines. Validation uses three complementary signals: computational similarity scores, zero-shot transfer performance of multilingual transformers (mBERT and XLM-R) on four NLP tasks (dependency parsing, named entity recognition, sentiment analysis, and abusive language identification) across eight languages, and an expert-linguist similarity survey. Across tasks and models, higher linguistic similarity is associated with better transfer, and the survey provides independent support for the computational metrics. Full article

Medical question answering (QA) systems and conversational agents have attracted growing interest as tools that can assist clinicians, support medical students, and help patients navigate complex information sources. However, existing evaluations of retrieval strategies largely overlook the cost–benefit balance—here referred to as frugality, under realistic computational constraints. This work introduces a frugality-based evaluation framework that jointly assesses accuracy improvements and computational cost to determine when retrieval-augmented generation is beneficial in medical question answering, rather than evaluating retrieval effectiveness through accuracy alone. This study addresses these gaps through a systematic comparative framework that evaluates retrieval relevance, computational efficiency, and knowledge base composition across multiple biomedical QA tasks. We employ open-source LLMs (LlaMA-3-8B-Instruct, Mistral-7B-Instruct-v0.3, and DeepSeek-7B-Chat) across three benchmark medical QA datasets, including MedMCQA, MedQA-USMLE, and PubMedQA. In addition to that, we evaluate a dataset with larger contexts to simulate model distraction across the CliniQG4QA dataset using additional models (Meditron-7B, Qwen2.5-7B-Medical, Medgemma-4B, Phi-3-mini-4k-Instruct, and GPT4o-Mini). We examine how retrieval design choices alter the accuracy–latency trade-off, examining how relevance, corpus design, and hardware constraints interact in medical retrieval-augmented generation (RAG) systems. Our comprehensive results demonstrate when retrieval is genuinely beneficial versus when it imposes unnecessary computational costs, highlighting interactions between relevance and corpus designs in QA. Full article

Mapping thematic structure in large scientific corpora enables the systematic analysis of research trends and conceptual organization. This work presents an unsupervised framework that leverages large language models (LLMs) as fixed semantic inference operators guided by structured soft prompts. The framework transforms raw abstracts into normalized semantic representations that reduce stylistic variability while retaining core conceptual content. These representations are embedded into a continuous vector space, where density-based clustering identifies latent research themes without predefining the number of topics. Cluster-level interpretation is performed using LLM-based semantic decoding to generate concise, human-readable descriptions of the discovered themes. Experiments on ICML and ACL 2025 abstracts demonstrate that the method produces coherent clusters reflecting problem formulations, methodological contributions, and empirical contexts. The findings indicate that prompt-driven semantic normalization combined with geometric analysis provides a scalable and model-agnostic approach for unsupervised thematic discovery across large scholarly corpora. Full article

Multiple Sclerosis (MS) and Myelitis are serious inflammatory spinal cord disorders with overlapping clinical symptoms and radiological characteristics, making accurate differentiation challenging yet clinically essential. Early and precise diagnosis is critical for guiding treatment strategies and improving patient outcomes. In this study, we propose KhayyamNet, a novel hybrid deep learning architecture designed to fuse complementary local and global representations for the accurate diagnosis of MS and Myelitis using spinal MRI. To improve robustness and generalization capability, a comprehensive preprocessing strategy including data augmentation and intensity normalization is also applied to reduce noise and address data variability. The proposed architecture combines three complementary deep learning models for feature extraction composed of Xception for high-level semantic features, Convolutional Neural Networks (CNNs) for fine-grained local patterns, and Vision Transformers (ViTs) for global contextual representations via attention mechanisms. Extracted features are then fused and refined using the Minimum Redundancy Maximum Relevance (MRMR) algorithm to eliminate redundancy and retain the most informative signals. Finally, a Random Forest (RF) classifier utilizes the optimized feature set to achieve accurate and robust differentiation between MS, Myelitis, and control spinal MRIs. Experimental results demonstrate that KhayyamNet outperforms existing methods by achieving an average classification accuracy of $98.15\pm 0.80\%$. This framework demonstrates promising performance for the automated analysis of spinal MRIs and shows potential to assist in the differentiation of MS and Myelitis. While these findings highlight the potential of KhayyamNet for automated MRI interpretation, its evaluation is limited to a single-center dataset, and further validation on external multi-center data is required. Full article

The creation and application of classification category labels are essential tasks for transforming complex information into structured knowledge. Categories are used for summary and reporting purposes and have historically been identified by domain experts based on their past experiences and norms. Our interest lies in the general case where expert-generated category lists require improvement, and unsupervised learning, on its own, struggles to effectively identify categories for multi-class classification of human-generated texts. We hypothesise that including an annotated knowledge graph (KG) in an embedding process will positively impact unsupervised clustering performance. Our goal is to identify clusters that can be labelled and used for classification. We look at unsupervised clustering of Maintenance Work Order (MWO) texts. MWOs capture vital observations about equipment failures in process and heavy industries. The selected KG contains a mapping of equipment types to their inherent function based on the IEC 81346-2 international standard for classification of objects in industrial systems. Performance is assessed by statistical analysis, subject matter experts, and Normalized Mutual Information score. We demonstrate that Word2Vec Bi-LSTM and Sentence-BERT NN embedding methods can leverage equipment inherent function information in the KG to improve failure mode cluster identification for the MWO. Organisations seeking to use AI to automate assignment of a failure mode code to each MWO currently need test sets classified by humans. The results of this work suggest that a semantic layer containing a knowledge graph mapping equipment types to inherent function, and inherent function to failure modes could assist in quality control for automated failure mode classification. Full article

Anomaly detection on tabular data is widely used in fraud detection, predictive maintenance, and medical screening. While heterogeneous ensembles combining multiple detection paradigms achieve strong performance, their computational cost limits deployment in latency-sensitive or resource-constrained environments. We propose KD-AnomalyNet, a teacher–student framework that distills anomaly knowledge from a high-capacity ensemble into a lightweight neural model for efficient inference. Beyond performance replication, we study how anomaly representations transfer during distillation. To this end, we introduce a noise perturbation analysis that serves as a diagnostic probe for representation stability without introducing additional trainable components. Experiments on ten benchmark datasets show that the distilled model preserves up to 98.5% of the teacher’s AUC-ROC on the nine capacity-sufficient datasets (84.7% mean retention across all ten datasets) while achieving 26–181× inference speedups. Our analysis reveals which forms of anomaly knowledge transfer reliably—global outliers (78% transfer) and isolation-based detection (88% retention)—and which degrade under compression—local outliers (20% transfer) and neighborhood-based detection (76% retention)—providing practical guidance for deploying distilled anomaly detectors. Full article

This work presents a fast neural surrogate capable of reconstructing fully three-dimensional hemodynamic velocity fields in stenotic and bifurcating microvascular geometries with satisfactory accuracy, avoiding repeated, computationally demanding computational fluid dynamics (CFD) simulations. A Fourier-augmented, coordinate-neural surrogate is presented and assessed for rapid computation of three-dimensional blood-flow fields in a sample geometry. The model is trained on detailed CFD data across a parameter set of stenosis severities that feed a direct mapping from spatial coordinates to velocity components. To mitigate spectral bias and improve accuracy in regions of steep gradients, the input space is embedded with random Fourier features and compared against a conventional multilayer perceptron (MLP) backbone. Predictive ability is assessed upon strict hold-out testing, during which certain arteriolar stenosis cases are excluded from training and treated with the Fourier surrogate. Direct comparison with CFD results reveals that the Fourier MLP achieves nearly CFD fidelity with the coefficient of determination R² ≥ 0.994 and offers more than 80% reduction in the normalized errors as provided by conventional MLP, with the precise improvement depending on the severity of stenosis. Centerline velocity and cross-sectional profiles further show that the Fourier MLP reconstructs stenosis speed-up and radial profiles more reliably compared to conventional MLP. These results indicate that Fourier feature embedding provides a simple and effective route to robust full-field hemodynamic surrogates for efficient screening of stenosis configurations without resorting to repeated, heavily demanding CFD simulations. Full article

Hit Song Science aims to forecast a song’s success before release and benefits from integrating signals beyond audio content alone. We present DeepHits, an end-to-end multimodal network that combines (i) log-Mel spectrogram embeddings from a compact residual 2D-CNN, (ii) frozen multilingual BERT lyric embeddings, and (iii) structured numeric features including high-level Spotify audio descriptors and contextual metadata (artist popularity, release year). Evaluated on 92,517 tracks from the SpotGenTrack dataset, DeepHits achieves a macro-F1 of 52.20% (accuracy 82.63%) in the established three-class setting and a macro-F1 of 23.15% (accuracy 37.00%) in a ten-class decile benchmark. To contextualize fine-grained performance, we report capacity-controlled shallow baselines, including metadata-only and early/late fusion variants, and show that the deep multimodal model provides a clear gain over these references (e.g., metadata-only: macro-F1 20.92%; accuracy 34.22%). Ablation results indicate that removing metadata yields the largest degradation in class-balanced performance, highlighting the strong predictive value of artist popularity and release year. Overall, DeepHits provides a reproducible benchmark and modality analysis for fine-grained popularity prediction under class imbalance. Full article

Recent benchmarks increasingly report that large language models (LLMs) exhibit human-like causal reasoning abilities, including counterfactual inference and intervention planning. However, many such evaluations rely on domains that are heavily represented in training data and embed strong semantic cues, raising the possibility that apparent causal competence may reflect semantic pattern recombination rather than structure-sensitive causal reasoning. Drawing on human developmental theories of causal induction, this perspective argues that genuine causal understanding requires robustness to novelty and reliance on conditional structure rather than semantic familiarity. To illustrate the testability of this claim, the paper includes a pilot demonstration using synthetic causal micro-worlds. Identical numerical evidence was presented to a LLM under two conditions: semantically meaningful variable labels and non-semantic coded labels. Across paired cases, the model reliably selected the correct causal structure when labels were meaningful, but frequently misidentified or exhibited instability in causal model selection under coded labels, despite producing locally coherent explanations. These divergences emerged most clearly in diagnostically ambiguous settings requiring suppression of misleading marginal associations. The results align with the claim that semantic scaffolding can support and stabilize apparent causal competence in LLMs without implying structure-sensitive reasoning. Full article

The growing demand for personalized human–computer interaction calls for methods that jointly model emotional states and personality traits. However, large-scale multimodal corpora annotated for both tasks are still lacking. This challenge stems from integrating diverse, task-specific corpora with divergent modality informativeness and domain characteristics. To address it, we propose SSL-MEPR, a semi-supervised multi-task cross-domain learning framework for Multimodal Emotion and Personality Recognition, which enables cross-task knowledge transfer without jointly labeled data. SSL-MEPR employs a three-stage strategy, progressively integrating unimodal single-task, unimodal multi-task, and multimodal multi-task models. Key innovations include Graph Attention Fusion, task-specific query-based cross-attention, predict projectors, and guide banks, which enable robust fusion and effective use of semi-labeled data via a modified GradNorm method. Evaluated on MOSEI (emotion) and FIv2 (personality), SSL-MEPR achieves a mean Weighted Accuracy (mWACC) of 70.26 and a mean Accuracy (mACC) of 92.88 in single-task cross-domain settings, outperforming state-of-the-art methods. Multi-task learning reveals domain-induced misalignment in modality informativeness but still uncovers consistent psychological patterns: sadness correlates with lower personality trait scores, while happiness aligns with higher ones. This work establishes a new paradigm for extracting cross-task psychological knowledge from disjoint multimodal corpora, demonstrating that semi-supervised multi-task cross-domain learning can bridge annotation gaps while preserving theoretically grounded emotion–personality relationships. Full article

Human Activity Recognition (HAR) on wearable and IoT devices must jointly satisfy four requirements: high accuracy, the ability to recognize previously unseen activities, strict memory and latency constraints, and interpretable decisions. In this work, we address all four by introducing an explainable Kolmogorov–Arnold Network for Human Activity Recognition (TinyKAN-HAR) with a zero-shot learning (ZSL) module, designed specifically for TinyML edge devices. The proposed KAN replaces fixed activation functions by learnable one-dimensional spline operators applied after linear mixing, yielding compact yet expressive feature extractors whose internal nonlinearities can be directly visualized. On top of the KAN latent space, we learn a semantic projection and cosine-based compatibility function that align sensor features with class-level semantic embeddings, enabling both pure and generalized zero-shot recognition of unseen activities. We evaluate our method on three benchmark datasets (UCI HAR, WISDM, PAMAP2) under subject-disjoint and zero-shot splits. TinyKAN-HAR consistently achieves over 97% macro-F1 on seen classes and over 96% accuracy on unseen activities, with harmonic mean above 96% in the generalized ZSL setting, outperforming CNN, LSTM and Transformer-based ZSL baselines. For explainability, we combine gradient-based attributions, SHAP-style global relevance scores and inspection of the learned spline functions to provide sensor-level, temporal and neuron-level insights into each prediction. After 8-bit quantization and TinyML-oriented optimizations, the deployed model occupies only 145 kB of flash and 26 kB of RAM, and achieves an average inference latency of 4.1 ms (about 0.32 mJ per window) on a Cortex-M4F-class microcontroller, while preserving accuracy within 0.2% of the full-precision model. These results demonstrate that explainable, zero-shot HAR with near state-of-the-art accuracy is feasible on severely resource-constrained TinyML edge devices. Full article

Modern multi-Unmanned Aerial Vehicle (UAV) attacks pose significant challenges to existing counter-UAV frameworks due to their agility, irregular spatial formations, and increasing reliance on intelligent evasive behaviors. This paper proposes a unified interception architecture that integrates Density-Based Spatial Clustering of Applications with Noise (DBSCAN) for multi-target grouping, a deceptive waypoint sequencing (DWS) mechanism for adversarial evasion, and a robust sliding-mode backstepping controller augmented with extended state observers (ESOs) for precise tracking under disturbances. DBSCAN enables real-time clustering of attacking UAVs without prior knowledge of the number of formations, producing dynamic centroids that serve as tactical interception references. To counter risky attackers capable of predicting defender trajectories, a novel DWS strategy introduces centroid-relative waypoints that preserve mission objectives while reducing trajectory predictability. Lyapunov-based analysis is developed for stability, guaranteeing uniform ultimate boundedness of the tracking errors. The proposed approach achieves successful interception in both scenarios, with an interception time of 7 s and final interception error of $0.023$ m in the single-UAV case, and an interception time of 8 s with final interception error of $0.050$ m in the multiple-UAV case, whereas the PID baseline fails to achieve interception under the same conditions. Extensive simulations involving single and multi-cluster engagements demonstrate that the proposed strategy achieves fast, accurate, and deception-resilient interception, outperforming the conventional PID approach in the presence of disturbances, nonlinearities, and dynamic swarm configurations. The obtained results show the effectiveness of integrating adaptive clustering, deceptive planning, and robust nonlinear control for modern UAV–UAV defensive operations. Full article