Search Results (2,916)

The proliferation of Internet of Things (IoT) devices has created unprecedented security challenges characterized by resource constraints, heterogeneous network architectures, and severe class imbalance in attack detection datasets. This paper presents a comprehensive benchmark evaluation of five Generative Adversarial Network (GAN) architectures for energy-aware intrusion detection: Standard GAN, Progressive GAN (PGAN), Conditional GAN (cGAN), Graph-based GAN (GraphGAN), and Wasserstein GAN with Gradient Penalty (WGAN-GP). Our evaluation framework introduces novel energy-normalized performance metrics, including Accuracy-per-Joule (APJ) and F1-per-Joule (F1PJ), that enable principled architecture selection for energy-constrained deployments. We propose an optimized WGAN-GP architecture incorporating diversity loss, feature matching, and noise injection mechanisms specifically designed for classification-oriented data augmentation. Experimental results on a stratified subset of the BoT-IoT dataset (approximately 1.83 million records) demonstrate that our optimized WGAN-GP achieves state-of-the-art performance, with 99.99% classification accuracy, a 0.99 macro-F1 score, and superior generation quality (MSE 0.01). While traditional classifiers augmented with SMOTE (i.e., Logistic Regression and CNN1D-TCN) also achieve 99.99% accuracy, they suffer from poor minority class detection (77.78–80.00%); our WGAN-GP improves minority class detection to 100.00% on the reported test split (45 of 45 attack instances correctly identified). Furthermore, WGAN-GP provides substantial efficiency advantages under our energy-normalized metrics, achieving superior accuracy-per-joule performance compared to Standard GAN. Also, a cross-dataset validation across five benchmarks (BoT-IoT, CICIoT2023, ToN-IoT, UNSW-NB15, CIC-IDS2017) was implemented using 250 pooled test attacks to confirm generalizability, with WGAN-GP achieving 98.40% minority class accuracy (246/250 attacks detected) compared to 76.80% for Classical + SMOTE methods, a statistically significant 21.60 percentage point improvement ($p<0.0001$). Finally, our analysis reveals that incorporating diversity-promoting mechanisms in GAN training simultaneously achieves best generation quality AND best classification performance, demonstrating that these objectives are complementary rather than competing. Full article

Spermatogenesis is a tightly coordinated differentiation program that sustains male fertility while transmitting genetic and epigenetic information to the next generation. This review consolidates mechanistic evidence showing how RNA-centered regulation integrates with the epitranscriptome and three-dimensional (3D) genome architecture to orchestrate germ-cell fate transitions from spermatogonial stem cells through meiosis and spermiogenesis. Recent literature is critically surveyed and synthesized, with particular emphasis on human and primate data and on stage-resolved maps generated by single-cell and multi-omics technologies. Collectively, available studies support a layered regulatory model in which RNA-binding proteins and RNA modifications coordinate transcript processing, storage, translation, and decay; small and long noncoding RNAs shape post-transcriptional programs and transposon defense; and dynamic chromatin remodeling and 3D reconfiguration align transcriptional competence with recombination, sex-chromosome silencing, and genome packaging. Convergent nodes implicated in spermatogenic failure are highlighted, including defects in RNA metabolism, piRNA pathway integrity, epigenetic reprogramming, and nuclear architecture, and the potential of these frameworks to refine molecular phenotyping in male infertility is discussed. Finally, key gaps and priorities for causal testing in spatially informed, stage-specific experimental systems are outlined. Full article

Multi-branch deep neural networks like AASIST3 achieve state-of-the-art comparable performance in audio anti-spoofing, yet their internal decision dynamics remain opaque compared to traditional input-level saliency methods. While existing interpretability efforts largely focus on visualizing input artifacts, the way individual architectural branches cooperate or compete under different spoofing attacks is not well characterized. This paper develops a framework for interpreting AASIST3 at the component level. Intermediate activations from fourteen branches and global attention modules are modeled with covariance operators whose leading eigenvalues form low-dimensional spectral signatures. These signatures train a CatBoost meta-classifier to generate TreeSHAP-based branch attributions, which we convert into normalized contribution shares and confidence scores ($C_b$) to quantify the model’s operational strategy. By analyzing 13 spoofing attacks from the ASVspoof 2019 benchmark, we identify four operational archetypes—ranging from “Effective Specialization” (e.g., A09, Equal Error Rate (EER) 0.04%, $C=1.56$) to “Ineffective Consensus” (e.g., A08, EER 3.14%, $C=0.33$). Crucially, our analysis exposes a “Flawed Specialization” mode where the model places high confidence in an incorrect branch, leading to severe performance degradation for attacks A17 and A18 (EER 14.26% and 28.63%, respectively). These quantitative findings link internal architectural strategy directly to empirical reliability, highlighting specific structural dependencies that standard performance metrics overlook. Full article

Accurate prediction of individual treatment effects in survival analysis is often complicated by the presence of competing risks and the inherent unobservability of counterfactual outcomes. While machine learning models offer improved discriminative power, they typically lack rigorous guarantees for uncertainty quantification, which are essential for safety-critical clinical decision-making. In this paper, we introduce CONFIDE (CONFormal Inference for Distribution-free Estimation), a novel framework that bridges causal inference and conformal prediction to construct valid prediction sets for cause-specific cumulative incidence functions. Unlike traditional confidence intervals for population-level parameters, CONFIDE provides individual-level prediction sets for time-to-event outcomes, which are more clinically actionable for personalized treatment decisions by directly quantifying uncertainty in future patient outcomes rather than uncertainty in population averages. By integrating semi-parametric hazard estimation with targeted bias correction strategies, CONFIDE generates calibrated prediction sets that cover the true potential outcome with a user-specified probability, irrespective of the underlying data distribution. We empirically validate our approach on four diverse medical datasets, demonstrating that CONFIDE achieves competitive discrimination (C-index up to 0.83) while providing robust finite-sample marginal coverage guarantees (e.g., 85.7% coverage on the Bone Marrow Transplant dataset). We note two key limitations: (1) coverage may degrade under heavy censoring (>40%) unless inverse probability of censoring weighted (IPCW) conformal quantiles are used, as demonstrated in our sensitivity analysis; (2) while the method guarantees marginal coverage averaged over the covariate distribution, conditional coverage for specific covariate values is theoretically impossible without structural assumptions, though practical approximations via locally-adaptive calibration can improve conditional performance. Our framework effectively enables trustworthy personalized risk assessment in complex survival settings. Full article

Immersive virtual reality is increasingly used for safety training, yet many initiatives remain technology-led pilots that enhance scenario realism and engagement without explaining how training becomes embedded in everyday work (e.g., alignment with SOPs, assessment routines, scheduling, and accountable debrief practices) or how skills reliably transfer back to duty. This paper addresses that gap by introducing Work-Integrated Safety Training (WIST) as a socio-technical training approach that couples IVR-based immersion with work-integrated routines to develop competence in safety-critical, passenger-facing work. Using Action Design Research (ADR) with Sweden’s national rail operator (SJ), we iteratively designed and evaluated a gamified immersive prototype for onboard conflict management, drawing on interviews, incident reports, co-design workshops, and in situ evaluations. We formalize four transferable design principles—specified with mechanisms and boundary conditions that guide how immersive training can (i) scaffold composure before intervention, (ii) make dynamic risk legible through interpretable cues, (iii) support SOP-aligned adaptive communication with replay-based reflection, and (iv) strengthen team coordination through role-specific checkpoints and psychologically safe debriefs. The paper contributes design knowledge for moving from isolated IVR demonstrations to work-integrated training systems that are implementable in organizations and testable in further longitudinal evaluation. Full article

(1) Background: Information and Communication Technology (ICT) integration in secondary education remains a critical challenge despite substantial investments in teacher training and infrastructure. This study investigated ICT certification levels, implementation patterns, and barriers among Greek secondary school teachers to understand the disconnect between policy aspirations and classroom realities. (2) Methods: A quantitative cross-sectional survey design was employed with 108 secondary teachers (61.1% female; mean age 47.3 years; 70.4% with >10 years’ experience) in the Prefecture of Ilia, Greece (response rate: 87.7%). Participants were permanent secondary school teachers employed in public schools during the 2021–2022 academic year; substitute teachers and private school staff were excluded. A three-section structured questionnaire was developed through literature review, expert validation (n = 3), and pilot testing (n = 10). Section A assessed demographics (5 items), Section B measured perceived barriers using a 7-item Likert scale, and Section C assessed implementation practices using a 10-item frequency scale (Cronbach’s α = 0.942). Data were analyzed using descriptive statistics, Mann–Whitney U tests, Kruskal–Wallis tests, and correlation analyses. (3) Results: While 74.1% of teachers held Level A certification, only 25.9% achieved Level B, with overall implementation remaining moderate (M = 2.92/5.00). Leadership support deficiency emerged as the primary barrier (76.9%), followed by inadequate technical support (74.1%). Younger teachers (24–35 years) demonstrated significantly higher ICT implementation than their older colleagues (56+ years), and teachers with less experience showed greater implementation frequency than veteran teachers—a finding that paradoxically challenges the “digital natives” assumption, given the barriers they face. Teachers preferred flexible Internet resources to formal educational software, indicating strategic adaptation to institutional constraints. Key limitations include convenience sampling, cross-sectional design, self-reported measures, and regional specificity. (4) Conclusions: The certification–implementation gap reveals that individual competencies cannot overcome unsupportive institutional environments. Effective ICT integration requires systemic transformation, encompassing leadership development, technical support, and structural reforms beyond traditional teacher training approaches. Full article

African Swine Fever is a lethal hemorrhagic disease caused by a DNA virus that affects domestic and wild pigs, causing serious economic losses in the swine industry. African Swine Fever virus (ASFV) is maintained in a sylvatic cycle that includes wildlife and Ornithodoros tick species. A huge investigation about ASFV structure and its infection process in pigs has been carried out in recent years, and although these studies have increased our knowledge about its pathogenesis, there are still many unclear aspects about which immune responses protect swine hosts against the disease caused by this virus. The mechanisms of ASFV infection in ticks are even less well understood. This infection is long term and persistent, with relatively high levels of virus replication in different tick tissues. According to specific infected tissues, the Ornithodoros tick species that are ASFV-competent vectors show transstadial, transovarial and/or venereal transmissions. This review is focused on the main process taking place at the virus–vector interface, summarizing the latest findings about the molecular and cellular aspects of ASFV infection in ticks, which could constitute the basis for developing novel strategies to interrupt the arthropod transmission cycle. Full article

This study explores the integration of technology and artificial intelligence (AI) as catalysts for professional teacher development within the context of Mexico’s educational challenges. Adopting a qualitative and exploratory approach, a four-phase workshop was conducted with 40 high school and university-level teachers from the National Polytechnic Institute (IPN). The methodology included scientific modeling activities using traditional methods, software (Tracker, ver. 6.2.0), and AI tools (ChatGPT-3.5), while analyzing participants’ perceptions and experiences. The findings reveal a clear disconnect between teachers’ theoretical competencies and their practical skills, with persistent gaps in scientific literacy at both educational levels. However, this study documents that the workshop functioned as a genuine professional learning community, where inter-academic collaboration and peer-learning proved to be an effective strategy for addressing these deficiencies. Technology, specifically the Tracker software, served as a catalyst for conceptual understanding. Despite AI’s potential for research, its limitations in the precision of responses reinforced this study’s central conclusion: technology does not replace the teacher’s work but transforms the teacher’s role into a critical mediator, responsible for guiding students to develop analytical and critical thinking in a complex digital environment. Full article

Eight Pacific Island Countries and Territories (PICTs) have been validated by the World Health Organization (WHO) as having eliminated lymphatic filariasis (LF) as a public health problem. WHO recommends that these countries implement post-validation surveillance (PVS) to ensure resurgence has not occurred. Some PICTs proactively conducted LF PVS even in the absence of specific recommendations or best-practice guidelines at the time of implementation. We aimed to explore the barriers and facilitators to implementing LF PVS in PICTs, with a view to informing context-specific strategies and regional guidelines. Key informant interviews were held between March and September 2024 with 15 participants involved in LF and/or neglected tropical disease surveillance. Transcripts were analysed thematically using a generalised deductive approach. A conceptual framework was developed to summarise themes with two main streams of barriers identified. Stream One Barriers included limited awareness of, and guidelines for, PVS requirements and competing national health priorities. Stream Two Barriers included cost, resource, and logistical barriers to conducting PVS. Participants called for clearer, contextually tailored guidelines, improved communication from WHO, and integration within existing systems. This study highlights the urgent need for operational guidance, policy advocacy, and capacity strengthening to ensure sustainable LF PVS in PICTs. Incorporating local context and leveraging existing health structures will be essential to prevent disease resurgence and maintain gains achieved through elimination programmes. Full article

Modern 5G+grade low power wide area network (LPWAN) technologies such as Narrowband Internet-of-Things (NB-IoT) operate utilizing a multi-channel slotted ALOHA algorithm at the random access phase. As a result, the random access phase in such systems is characterized by relatively low throughput and is highly sensitive to traffic fluctuations that could lead the system outside of its stable operational regime. Although theoretical results specifying the optimal transmission probability that maximizes the successful preamble transmission probability are well known, the lack of knowledge about the current offered traffic load at the BS makes the problem of maintaining the optimal throughput a challenging task. In this paper, we propose and analyze a new reactive access-barring scheme for NB+IoT systems based on machine learning (ML) techniques. Specifically, we first demonstrate that knowing the number of user equipments (UE) experiencing a collision at the BS is sufficient to make conclusions about the current offered traffic load. Then, we show that through utilizing ML-based techniques, one can safely differentiate between events in the Physical Random Access Channel (PRACH) at the base station (BS) side based on only the signal-to-noise ratio (SNR). Finally, we mathematically characterize the delay experienced under the proposed reactive access-barring technique. In our numerical results, we show that by utilizing modern neural network approaches, such as the XGBoost classifier, one can precisely differentiate between events on the PRACH channel with accuracy reaching 0.98 and then associate it with the number of user equipment (UE) competing at the random access phase. Our simulation results show that the proposed approach can keep the successful preamble transmission probability constant at approximately 0.3 in overloaded conditions, when for conventional NB-IoT access, this value is less than 0.05. The proposed scheme achieves near-optimal throughput in multi-channel ALOHA by employing dynamic traffic awareness to adjust the non-unit transmission probability. This proactive congestion control ensures a controlled and bounded delay, preventing latency from exceeding the system’s maximum load capacity. Full article

Regulatory T cells (Tregs), particularly their phenotypically distinct subpopulations, are critical for the establishment of maternal immune tolerance during embryo implantation. Despite advances in assisted reproductive technologies, implantation failure remains a frequent and often unexplained clinical challenge. Variations in Treg frequency and phenotype have been proposed to influence implantation success, particularly under differing hormonal conditions. This study aimed to investigate peripheral blood Treg levels and their subpopulations on the day of blastocyst transfer in both stimulated in vitro fertilization (IVF/ICSI) cycles involving controlled ovarian hyperstimulation (COH) and true natural cycles with frozen embryo transfer (FET), and to examine their associations with systemic hormone levels and anti-Müllerian hormone (AMH). A prospective observational study was conducted including women undergoing IVF/ICSI with fresh embryo transfer (ET) and women undergoing natural cycle FET. Peripheral blood samples were collected on the day of ET and analyzed using 13-colour flow cytometry, enabling detailed subdivision of Tregs into multiple subpopulations based on the expression of differentiation and chemokine markers, including CXCR5. In addition, because common $\gamma$-chain cytokines may influence pregnancy success by modulating the balance between suppressive Treg and non-Treg subsets, intracellular STAT5 signaling was assessed using phospho-specific flow cytometry. Serum estradiol, progesterone, FSH, LH, and AMH levels were measured in parallel. Significant differences were observed in Treg subpopulation distributions between women who conceived and those who did not. Higher frequencies of naïve CXCR5⁻ Tregs were associated with clinical pregnancy, independent of age, and correlated with serum progesterone levels. Moreover, both naïve Treg frequency and enhanced IL-7-dependent STAT5 signaling in naïve Tregs from women undergoing COH were associated with AMH levels, suggesting a link between ovarian reserve and Treg homeostasis mediated by signal transducer and activator of transcription 5 (STAT5) signaling. In conclusion, Treg subpopulations, particularly CXCR5⁻ naïve Tregs, appear to play a central role in implantation success following ET. Their distribution differs between stimulated and natural cycles and is influenced by systemic progesterone levels and STAT5 signaling. These findings suggest that peripheral Treg profiling may represent a potential biomarker of implantation competence and could inform personalized approaches in assisted reproduction. Full article

Background: Nurses’ clinical reasoning is increasingly being recognized as a core competence that enhances the quality and safety of care across diverse healthcare settings. Nurses with refined clinical reasoning skills contribute significantly to improved health outcomes and broader health gains. In emergency departments, this competence is essential to rapidly assessing complex problems, anticipating complications, and preventing the deterioration of patients’ clinical conditions. Such expertise enables nurses to discern the severity of clinical situations quickly and intervene effectively. Objectives: The aims of this study were to analyze the clinical reasoning process of nurses and develop a theory that explains this process in emergency care settings. Methodology: This qualitative study explored the following research question: “How do nurses enact the clinical reasoning process in emergency departments?” The Grounded Theory methodology was used, with a theoretical sample of 20 nurses. Data collection methods included in-depth interviews, participant observation, and field notes. Results: The theoretical analysis identified clinical reasoning as a substantive theory composed of two subprocesses: Diagnostic Nursing Assessment and Therapeutic Nursing Intervention. Nurses’ clinical reasoning determines two levels of patient severity—Level I, life-threatening situations (immediate risk), and Level II, non-life-threatening situations (expressed problems)—according to which nursing interventions are adjusted. Conclusions: The Nursing Clinical Reasoning Model is a dynamic and continuous process that involves both Diagnostic Nursing Assessment and Nursing Therapeutic Intervention. It is deeply rooted in the nurse–patient–family relationship and is shaped by the specific care context, which influences nurses’ assessments and interventions and patients’ responses and behaviors. Full article

Effective communication in English is a critical behavioral competency for seafarers in a multilingual maritime environment, directly impacting operational safety. However, a gap exists between current Maritime English (ME) training in China and the actual communication demands of global seafaring. This study aims to identify the specific ME skills including linguistic, behavioral, and sociolinguistic dimensions that are most important for on-board performance and safety management from the perspective of pre-service maritime students at Shanghai Maritime University. A mixed-methods approach was used, combining structured questionnaires (n = 313) with in-depth follow-up interviews (n = 10). The results identified 24 highly needed ME skills, particularly focused on areas governing safety-critical behaviors, such as wireless communication, security protocols, and emergency procedures. In addition, based on learner profiling, the study depicts two different learner characteristics: exam-focused and work-focused students, each with different views on the importance of skills. Work-focused students place greater emphasis on the practicality of their skills. The interview data confirms and enriches these quantitative research results. The research findings emphasize that ME courses must be more closely aligned with real-world communicative scenarios and behaviors, prioritize scenario based teaching and practical operations, and tailor differentiated teaching based on learner psychology and behavioral preference. This study offers references for maritime education institutions with similar learner profiles to optimize ME curricula, prioritize secure communication skills, and strengthen industry-education collaboration, thereby enhancing pre-service maritime students’ safety behavior and professional competitiveness in China. Full article

Serious games are increasingly recognized as powerful pedagogical tools, often offering engaging, interactive, and practical learning experiences. This paper presents the design, implementation, and evaluation of a 3D virtual serious game specifically tailored for cybersecurity governance and policy education. In particular, the nature of the game is an escape room, drawing on military training principles: players must solve a problem to escape one room before advancing to the next. Set within a virtual company environment, the game features three interactive zones that guide students through analyzing cyber risks, aligning security frameworks, and drafting appropriate policies. This structure cultivates critical thinking and decision-making skills and strengthens practical cybersecurity competencies. The primary contribution lies in the integration of game-based learning and 3D virtual technology to create robust, hands-on educational materials. The design incorporates structural features that create barriers to generative AI delegation to address challenges related to generative AI misuse, ensuring that the activities cannot be easily replicated and thereby supporting academic integrity. A post-activity perception survey (n = 20) suggests that students found this approach both engaging and effective, with participants self-reporting enhanced understanding and enthusiasm toward cybersecurity governance and policy concepts. These findings highlight the potential of gamified environments to bridge theory and practice in cybersecurity education, equipping learners with industry-relevant skills while fostering deeper engagement and active learning. Full article

Weed management has become a critical agricultural practice, as weeds compete with crops for nutrients, host pests and diseases, and cause major economic losses. The invasive weed Ambrosia artemisiifolia (common ragweed) is particularly problematic in Hungary, endangering crop productivity and public health through its fast proliferation and allergenic pollen. This review examines the current challenges and impacts of A. artemisiifolia while exploring sustainable approaches to its management through precision agriculture. Recent advancements in unmanned aerial vehicles (UAVs) equipped with advanced imaging systems, remote sensing, and artificial intelligence, particularly deep learning models such as convolutional neural networks (CNNs) and Support Vector Machines (SVMs), enable accurate detection, mapping, and classification of weed infestations. These technologies facilitate site-specific weed management (SSWM) by optimizing herbicide application, reducing chemical inputs, and minimizing environmental impacts. The results of recent studies demonstrate the high potential of UAV-based monitoring for real-time, data-driven weed management. The review concludes that integrating UAV and AI technologies into weed management offers a sustainable, cost-effective, mitigate the socioeconomic impacts and environmentally responsible solution, emphasizing the need for collaboration between agricultural researchers and technology developers to enhance precision agriculture practices in Hungary. Full article