Search Results (626)

Accurate diagnosis of wheat leaf diseases in real farming conditions requires models that are not only highly accurate but also computationally efficient and interpretable for practical deployment on edge devices. We propose GateProtoNet (GPN), a two-stage, compute-aware, and explainable framework for multi-class leaf disease recognition. Stage-1 performs ultra-light healthy-versus-diseased screening, enabling early exit for healthy samples and substantially reducing average expected inference cost. For diseased samples, Stage-2 applies a novel hybrid backbone featuring a frequency-factorized Discrete Wavelet Transform (DWT) stem, parallel micro-lesion convolutional encoding for fine texture patterns, and a linear token mixer for global context modeling. A cross-gated fusion module adaptively integrates local and global evidence with minimal computational overhead. To ensure trustworthy predictions, GPN introduces a prototype evidence head that performs classification via similarity to learned class prototypes, providing human-interpretable explanations, along with a faithfulness constraint that enforces explanation reliability by measuring confidence degradation under salient region removal. Rigorous evaluation on four publicly available wheat and cotton leaf disease datasets demonstrate that GateProtoNet achieves 99.2% classification accuracy, 99.1% macro-F1 score, and 99.3% AUC, significantly outperforming existing CNN, transformer, and hybrid baselines while requiring substantially fewer parameters and FLOPs. The two-stage inference strategy reduces average computational cost by avoiding full model execution on healthy leaves, enabling real-time, on-device diagnosis for resource-constrained agricultural environments. Full article

Anomaly detection is crucial in maintaining the safety, reliability, and optimal performance of complex systems across diverse domains, such as industrial manufacturing, cybersecurity, and autonomous systems. While conventional methods typically handle single data modalities, recently, there has been an increase in the application of multimodal detection in dynamic real-world environments. This paper presents a comprehensive review of recent research at the intersection of agentic artificial intelligence and large language-based multimodal anomaly detection. We systematically analyze and categorize existing studies based on the agent architecture, reasoning capabilities, tool integration, and modality scope. The main contribution of this work is a novel taxonomy that unifies agentic and multimodal anomaly detection methods, alongside benchmark datasets, evaluation methods, key challenges, and mitigation strategies. Furthermore, we identify major open issues, including data alignment, scalability, reliability, explainability, and evaluation standardization. Finally, we outline future research directions, with a particular emphasis on trustworthy autonomous agents, efficient multimodal fusion, human-in-the-loop systems, and real-world deployment in safety-critical applications. Full article

Incentive-based demand response (DR) programs rely on accurate and trustworthy quantification of customer performance to ensure fair compensation and market efficiency. Estimating the customer baseline load is an important part of this process. It shows how much electricity would be used if there were no DR occurrence. Unlike conventional load forecasting, baseline modeling is inherently unobservable, economically sensitive, and vulnerable to strategic manipulation. With the growing penetration of distributed energy resources, electric vehicles, and intelligent control technologies, traditional baseline estimation approaches face increasing limitations. This paper offers a thorough and future-oriented synthesis of baseline load estimation for incentive-based DR strategies. Current approaches are carefully classified into rule-based, statistical, probabilistic, machine learning (ML), and hybrid intelligence techniques, and their appropriateness for various DR services and client categories is rigorously evaluated. Beyond modeling accuracy, this paper emphasizes market-oriented requirements, including incentive compatibility, simplicity, transparency, privacy preservation, and deployment feasibility. Furthermore, emerging digital trust enablers such as blockchain and FL are reviewed, along with baseline-free and baseline-light alternatives for performance evaluation. Finally, open research challenges and future directions toward interpretable, robust, and market-ready baseline intelligence are discussed. Full article

The advancement of generative artificial intelligence (GAI) in healthcare is driving a transition from dialogue-based medical chatbots to workflow-embedded clinical AI agents. These agentic systems incorporate persistent state management, coordinated tool invocation, and bounded autonomy, enabling multi-step reasoning within institutional processes. As a result, traditional response-level evaluation frameworks are insufficient for understanding system behavior. This review provides a conceptual synthesis of the evolution from conversational systems to agentic architectures and proposes a system-level modeling framework for ethical clinical AI agents. We identify core architectural dimensions, including autonomy gradients, state persistence, tool orchestration, workflow coupling, and human–AI co-agency, and examine how these features reshape bias propagation pathways, error cascade dynamics, trust calibration, and accountability structures. Emphasizing that ethical risks emerge from longitudinal system interactions rather than isolated outputs, we argue for embedding fairness constraints, transparency mechanisms, and lifecycle governance directly within AI design. By outlining trajectory-level evaluation strategies, equity-aware development approaches, collaborative oversight models, and adaptive regulatory frameworks, this paper establishes a foundation for the responsible and trustworthy integration of agentic AI in healthcare. Full article

Traditional intrusion detection systems for IoT networks achieve high classification accuracy but lack interpretability and actionable incident-response capabilities, limiting their operational value in security-critical environments. This paper presents a graph-based multi-agent framework that integrates ensemble machine learning with Large Language Model (LLM)-powered incident report generation via Retrieval-Augmented Generation (RAG). The system employs a three-phase architecture: (1) a lightweight Random Forest binary pre-detection, achieving 99.49% accuracy with a 6 MB model size for edge deployment; (2) ensemble classification combining Multi-Layer Perceptron, Random Forest, and XGBoost with soft voting and SHAP-based feature attribution for explainability; and (3) a ReAct-based summary agent that synthesizes classification results with external threat intelligence from Web search and scholarly databases to generate evidence-grounded incident reports. To address the challenge of evaluating non-deterministic LLM outputs, we introduce custom RAG evaluation metrics—faithfulness and groundedness implemented via the LLM-as-Judge framework. Experimental validation on the ACI IoT Network Dataset 2023 demonstrates ensemble accuracy exceeding 99.8% across 11 attack classes; perfect groundedness scores (1.0), indicating all generated claims derive from the retrieved context; and moderate faithfulness (0.64), reflecting appropriate analytical synthesis. The ensemble approach mitigates individual model weaknesses, improving the UDP Flood F1 score from 48% (MLP alone) to 95% through soft voting. This work bridges the gap between high-accuracy detection and trustworthy, actionable security analysis for automated incident-response systems. Full article

With the widespread adoption of anonymity networks such as Tor, I2P, and JonDonym, reliably classifying darknet traffic remains challenging due to feature redundancy and severe class imbalance in encrypted flows. Existing approaches often rely on static feature-selection strategies and generic oversampling methods, which limit robustness and may distort traffic semantics. This study proposes an adaptive classification framework integrating Adaptive Weighted Feature Aggregation (AWFA) for reliability-aware feature selection and Traffic-Aware SMOTE (TA-SMOTE) for semantically constrained perturbations of packet-size and timing features while preserving flow-level structure. The framework is evaluated on a two-layer hierarchy comprising browser-level (L1) and application-level (L2) classification. At the L2, the proposed AWFA and TA-SMOTE pipeline attains a macro-F1 score of 73.81%, significantly exceeding PCA-based reduction and traditional RF-based selection with SMOTE. At the browser level (L1), macro-F1 rises from 91.58% to 96.09% while reducing the feature space from 84 to 40 attributes, highlighting both performance improvements and structural efficiency gains. Additional semantic validation confirms that the balancing process preserves the statistical and structural characteristics of genuine darknet traffic. These results indicate that reliability-aware feature aggregation and traffic-aware balancing provide a practical, trustworthy approach to modern darknet traffic classification. Full article

Reinforcement learning (RL) agents are increasingly deployed for voltage control in power distribution networks. However, their opaque decision-making creates a significant trust barrier, limiting their adoption in safety-sensitive operational settings. This paper presents XRL-LLM, a novel framework that generates natural language explanations for RL control decisions by combining game-theoretic feature attribution (KernelSHAP) with large language model (LLM) reasoning grounded in power systems domain knowledge. We deployed a Proximal Policy Optimization (PPO) agent on an IEEE 33-bus network to coordinate capacitor banks and on-load tap changers, successfully reducing voltage violations by 90.5% across diverse loading conditions. To make these decisions interpretable, KernelSHAP identifies the most influential state features. These features are then processed by a domain-context-engineered LLM prompt that explicitly encodes network topology, device specifications, and ANSI C84.1 voltage limits.Evaluated via G-Eval across 30 scenarios, XRL-LLM achieves an explanation quality score of 4.13/5. This represents a 33.7% improvement over template-based generation and a 67.9% improvement over raw SHAP outputs, delivering statistically significant gains in accuracy, actionability, and completeness ($p<0.001$, Cohen’s d values up to 4.07). Additionally, a physics-grounded counterfactual verification procedure, which perturbs the underlying power flow model, confirms a causal faithfulness of 0.81 under critical loading. Finally, five ablation studies yield three broader insights. First, structured domain context engineering produces synergistic quality gains that exceed any single knowledge component, demonstrating that prompt composition matters more than the choice of foundational model. Second, even an open source 8B-parameter model outperforms templates given the same prompt, confirming the framework’s backbone-agnostic value. Most importantly, counterfactual faithfulness increases alongside load severity, indicating that post hoc attributions are most reliable in the high-stakes regimes where trustworthy explanations matter most. Full article

Background: Strength training (ST) is essential for enhancing athletic performance and reducing injury risk, yet traditional monitoring relies heavily on subjective assessment, limiting objective and individualized evaluation. Objective: This scoping review critically synthesizes current applications of artificial intelligence (AI) and wearable technologies (WT) in ST, with emphasis on methodological approaches, data characteristics, explainability, and practical readiness. Methods: Searches of PubMed and Scopus identified 13 peer-reviewed studies (2015–2025). Evidence was charted and synthesized to compare AI models, wearable sensor configurations, validation strategies, and translational potential. Results: Studies employed classical machine learning, deep learning, and hybrid approaches alongside inertial, force, strain, and physiological sensors to support exercise classification, load estimation, fatigue detection, and performance monitoring. Deep learning models dominated movement recognition tasks, whereas simpler models often aligned better with small datasets and interpretability requirements. However, most studies relied on limited, homogeneous samples and internal validation, restricting generalizability and real-world applicability. Explainability was inconsistently addressed, particularly in higher-risk applications such as injury prediction. Conclusions: AI-enhanced wearables provide objective and individualized ST monitoring, but current evidence remains largely experimental. To ensure a practical application is implemented, standardized datasets, robust external validation, and greater integration of explainable AI are required to support and deliver trustworthy decision-making. Full article

The Automated Essay Scoring (AES) systems confront two fundamental challenges: opaque “black-box” decision-making that limits educator trust, and insufficient validation across linguistically diverse educational contexts. This study proposes IRT-AESF, an innovative framework that bridges educational measurement theory and artificial intelligence by integrating item response theory (IRT) with deep learning. The framework generates three theoretically grounded psychometric parameters: student ability, item difficulty, and item discrimination, which provide transparent and interpretable explanations for scoring decisions. We rigorously evaluated IRT-AESF through 5-fold cross-validation on three large-scale datasets comprising 41,328 authentic essays from English and Chinese educational settings, including both classroom assessments and high-stakes examinations. Results demonstrate statistically significant improvements over competitive baseline models, achieving an 8.4% relative increase in quadratic weighted kappa while maintaining robust cross-lingual performance. This research advances the development of transparent, trustworthy automated assessment systems that deliver not only scores but meaningful diagnostic insights for educational practice. Full article

The Internet of Things (IoT) plays an increasingly central role in healthcare by enabling continuous patient monitoring, remote diagnosis, and data-driven clinical decision-making through interconnected medical devices and sensing infrastructures. Despite these advances, IoT healthcare systems remain constrained by persistent challenges related to data privacy, computational efficiency, scalability, and regulatory compliance. Federated learning (FL) reduces reliance on centralised data aggregation but remains vulnerable to inference-based privacy risks, while edge-oriented approaches are limited by device heterogeneity and restricted computational and energy resources; the deployment of large language models (LLMs) further exacerbates concerns surrounding privacy exposure, communication overhead, and practical feasibility. This study introduces Trustworthy Intelligence (TI) as a guiding framework for privacy-preserving distributed intelligence in IoT healthcare, explicitly integrating predictive performance, privacy protection, and deployment-oriented system design. Within this framework, split learning (SL) is examined as a core architectural mechanism and extended to support split-aware LLM integration across heterogeneous devices, supported by a structured taxonomy spanning architectural configurations, system adaptation strategies, and evaluation considerations. The study establishes a systematic mapping between SL design choices and representative healthcare scenarios, including wearable monitoring, multi-modal data fusion, clinical text analytics, and cross-institutional collaboration, and analyses key technical challenges such as activation-level privacy leakage, early-round vulnerability, reconstruction risks, and communication–computation trade-offs. An energy- and resource-aware adaptive cut layer selection strategy is outlined to support efficient deployment across devices with varying capabilities. A proof-of-concept experimental evaluation compares the proposed SL–LLM framework with centralised learning (CL), federated learning (FL), and conventional SL in terms of training latency, communication overhead, model accuracy, and privacy exposure under realistic IoT constraints, providing system-level evidence for the applicability of the TI framework in distributed healthcare environments and outlining directions for clinically viable and regulation-aligned IoT healthcare intelligence. Full article

Dry-type electrical transformers are essential components in commercial, industrial, and residential power distribution systems, as they adapt voltage levels required by a broad range of load types. Although they are robustly constructed, they are exposed to adverse operational and environmental conditions such as dust, humidity, and electrical disturbances that may cause premature winding damage, such as inter-turn short circuits. This study focuses on the detection of inter-turn short-circuit faults in a 15 kVA commercial dry-type transformer, where a fault equivalent to 11.54% of short-circuited turns was induced in the tap changers. Axial, radial, and rotational leakage magnetic flux signals were captured using a low-cost, non-invasive triaxial Hall-effect magnetic flux sensor. During data processing, Fisher Score feature selection was applied to identify the most relevant indicators. Subsequently, feature extraction techniques, including Linear Discriminant Analysis, Principal Component Analysis (PCA), Uniform Manifold Approximation and Projection, and Isometric Mapping, were evaluated. The technique that best preserved global and local data structures was selected using Trustworthiness, Spearman’s correlation, and Kruskal’s stress metrics. PCA was selected as the optimal technique based on these quality metrics, achieving the highest classification performance. The resulting subspace data were classified using support vector machines and applying K-fold cross-validation. The proposed system achieved classification accuracies above 95%, with high recall and F1-score values, for inter-turn fault detection in each winding, confirming its effectiveness for reliable inter-turn fault detection in each transformer winding. Full article

Deepfakes and adjacent synthetic-media capabilities have become a systemic challenge for information integrity, security, and digital trust. Countermeasures now span passive detection methods that infer manipulation from content traces, active provenance systems that cryptographically bind metadata to media, and watermarking approaches that embed detectable signals into content or generative processes. This review presents a rigorous synthesis of tools and technologies to combat deepfakes across modalities (image, video, audio, and selected multimodal settings), drawing primarily from the peer-reviewed literature, standardized benchmarks, and official technical specifications and reports. The review analyzes detection methods, provenance and authentication technologies, with emphasis on cryptographic manifests and threat models, watermarking and content provenance, including diffusion-era watermarking and industrial deployments, adversarial robustness and attacker adaptation, datasets and benchmarks, evaluation metrics across tasks, and deployment and scalability constraints. A dedicated section addresses legal, ethical, and policy issues, focusing on emerging transparency obligations and platform governance. The review finds that no single countermeasure is sufficient in realistic adversarial settings. The strongest practical approach is a layered defense that combines provenance, watermarking, content-based detection, and human oversight. The study concludes with limitations of the current evidence base and prioritized research directions to improve generalization, interoperability, and trustworthy user experiences. Full article

Major Depressive Disorder (MDD) is a significant global health burden that contributes to disability and reduced quality of life. Its impact extends beyond individuals, placing emotional, social, and economic strain on families and healthcare systems worldwide. Despite its prevalence, MDD remains widely misunderstood, with limited mental health literacy and persistent stigma often preventing individuals from seeking help. This research explored the prediction of MDD utilising social media data via Natural Language Processing (NLP), Machine Learning (ML), and explainable Machine Learning (xML) techniques. The research aimed at identifying depressive indicators on X (formerly Twitter) and developing interpretable models for depression risk detection. The study’s methodology followed the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework to ensure a systematic approach to data analysis. Data was collected via X’s API and processed using regex-based noise removal, normalisation, tokenisation, and lemmatisation. Symptoms were mapped to DSM-5-TR criteria at the post-level, with user-level MDD risk assessed based on symptom persistence over a two-week period. Risk levels were classified as No Risk, Monitor, and High Risk to facilitate early intervention. Six ML models were trained and tested, while the Synthetic Minority Over-sampling Technique (SMOTE) was applied to mitigate class imbalance. The dataset was partitioned into training and testing sets using an 80:20 split. ML models were evaluated, and the Extreme Gradient Boosting model outperformed the others. Extreme Gradient Boosting achieved an accuracy of 0.979, F1-score of 0.970, and ROC-AUC of 0.996, surpassing benchmark results reported in prior studies. Explainability techniques, such as LIME and tree-based feature importance, enhance model transparency and clinical interpretability. Depressed mood consistently emerged as the highest-weighted predictor across different models. The findings highlight the value of aligning ML models with validated diagnostic frameworks to improve trustworthiness and reduce false positives. Future research can expand beyond text-based analysis by incorporating multimodal features to broaden diagnostic depth. Full article

Quantifying predictive uncertainty is essential for safe and trustworthy real-world AI deployment. However, the fully nonparametric estimation of conditional distributions remains challenging for multivariate targets. We propose Tomographic Quantile Forests (TQF), a nonparametric, uncertainty-aware, tree-based regression model for multivariate targets. TQF learns conditional quantiles of directional projections ${\mathbf{n}}^{\top }\mathbf{y}$ as functions of the input $\mathbf{x}$ and the direction n. At inference, it aggregates quantiles across many directions and reconstructs the multivariate conditional distribution by minimizing the sliced Wasserstein distance via an efficient alternating scheme with convex subproblems. Unlike classical directional-quantile approaches that typically produce only convex quantile regions and require training separate models for different directions, TQF covers all directions with a single model to reconstruct the full conditional distribution itself, naturally overcoming any convexity restrictions. We evaluate TQF on synthetic and real-world datasets, and release the source code on GitHub. Full article

Growing consumer awareness of environmental issues has led to heightened interest in environmentally responsible personal care products, with notable growth in the green beauty market. Despite this trend, consumer evaluations and purchasing decisions are shaped by numerous elements, particularly the trustworthiness of environmental information provided by brands. While elements such as eco-labels, environmentally friendly packaging, and individual concern for the environment are generally associated with favourable views of green marketing, they can also trigger doubt among consumers. This study investigates how eco-labeling, green packaging, and environmental concern influence consumer attitudes toward green beauty products and how these attitudes affect purchase intentions for them. The study included 500 South African consumer participants, and data were collected through an accredited research organization. The results indicate that eco-labeling, green packaging, and environmental concern each have a significant positive effect on consumer attitudes toward green beauty products. In addition, consumer attitudes were found to strongly influence purchase intentions among South African consumers. The study contributes to the literature by emphasizing the complex role of informational cues and potential perceptual barriers in sustainable consumption of green beauty products. It also provides practical insights for marketers, highlighting the importance of enhancing the credibility of eco-labels and reducing consumer doubt through clear, transparent communication strategies in the South African context. Full article