Search Results (8,201)

With the rapid development of the Internet of Things (IoT) and Industrial IoT (IIoT), Network Intrusion Detection Systems (NIDSs) play a critical role in securing modern networked environments. Despite advances in multi-class intrusion detection, existing approaches face challenges from high-dimensional heterogeneous traffic data, severe class imbalance, and limited interpretability of high-performance “black-box” models. To address these issues, this study presents an XGBoost-based NIDSs integrating optimized strategies for feature dimensionality reduction and class balancing, alongside SHAP-based interpretability analysis. Feature reduction is investigated by comparing selection methods that preserve original features with generation methods that create transformed features, aiming to balance detection performance and computational efficiency. Class balancing techniques are evaluated to improve minority-class detection, particularly reducing false negatives for rare attack types. SHAP analysis reveals the model’s decision process and key feature contributions. The experimental results demonstrate that the method enhances multi-class detection performance while providing interpretability and computational efficiency, highlighting its potential for practical deployment in IoT security scenarios. Full article

With the rapid advancement of clean energy, wind farm planning and construction are expanding worldwide, increasing the demand for accurate resource assessments. This study investigates the influence of vertical wind characteristics on wind farm siting and energy production, using measured meteorological data from the Hangjinqi wind farm. Results show that both mean wind speed increase substantially with altitude, indicating that upper layers provide richer and more stable wind resources. The estimated annual energy production of the site reaches 23,500 MWh, with capacity factors ranging from 35% to 42%, well above the national average. Seasonal and diurnal variations are evident: wind speeds strengthen during winter and spring, particularly at night, while turbulence intensity peaks in the daytime and decreases with height. Carbon dioxide (CO₂) reduction also increases with hub height, with the most pronounced seasonal reductions in spring (3367.6–5041.1 tCO₂, +49.7%) and winter (3215.7–5380.0 tCO₂, +67.4%), equivalent to several thousand tons of standard coal per turbine annually. Optimal performance is observed at 100–140 m, demonstrating efficient utilization of mid- to high-altitude resources. Nevertheless, discrepancies in turbine performance at different hub heights suggest untapped potential at higher elevations. These findings highlight the importance of incorporating vertical wind characteristics into wind farm siting decisions, and support the deployment of turbines with tower heights ≥140 m alongside intelligent scheduling and forecasting strategies to maximize energy yield and economic benefits. Full article

New ventures operating in volatile and unpredictable environments must rely on rapid adaptation and decisive action, making improvisation a critical entrepreneurial capability. This study examines how improvisation enhances new venture performance by uncovering the psychological and learning-based mechanisms through which its effects unfold. Drawing on the Knowledge-Based View (KBV) and Social Learning Theory (SLT), the model proposes that improvisation strengthens entrepreneurial self-efficacy, enabling entrepreneurs to approach uncertainty with greater confidence and adaptive judgment. Using a two-wave survey of 322 startup founders in Turkey and analyses conducted through PROCESS and complementary SEM estimation, the findings show that improvisation significantly boosts both entrepreneurial self-efficacy and new venture performance. Entrepreneurial self-efficacy emerges as a key mediating mechanism, indicating that improvisational experiences help entrepreneurs develop mastery, reinforce capability beliefs, and translate spontaneous action into improved outcomes. The results further suggest that improvisational episodes provide immediate learning cues that enhance situational awareness and decision-making agility, deepening the psychological pathway that links spontaneous behavior to venture performance. Additionally, relative explorative learning significantly moderates the relationship between improvisation and entrepreneurial self-efficacy, demonstrating that entrepreneurs benefit more from improvisation when they actively pursue new knowledge, experiment with unfamiliar approaches, and challenge routine assumptions. This moderating role clarifies when improvisation produces its strongest effects, while the mediating mechanism explains how performance improvements materialize through confidence-building processes. By integrating these mechanisms into a unified explanation, the study advances understanding of the improvisation–performance relationship and highlights the importance of learning-oriented behavior in converting spontaneous action into sustained entrepreneurial advantage. The findings offer theoretical contributions and actionable insights for entrepreneurs seeking to strengthen adaptability, resilience, and competitiveness in fast-changing environments. Full article

The increasing complexity of modern electrical substations—driven by renewable integration, advanced automation, and asset aging—necessitates a transition from reactive maintenance toward intelligent, data-driven strategies. Predictive maintenance (PdM), supported by artificial intelligence, enables early fault detection and remaining useful life (RUL) estimation, while Digital Twin (DT) technology provides synchronized cyber–physical representations for situational awareness and risk-free validation of maintenance decisions. This study proposes a five-layer DT-enabled PdM architecture integrating standards-based data acquisition, semantic interoperability (IEC 61850, CIM, and OPC UA Part 17), hybrid AI analytics, and cyber-secure decision support aligned with IEC 62443. The framework is validated using utility-grade operational data from the SS1 substation of the Badra Oil Field, comprising approximately one million multivariate time-stamped measurements and 139 confirmed fault events across transformer, feeder, and environmental monitoring systems. Fault detection is formulated as a binary classification task using event-window alignment to the 1 min SCADA timeline, preserving realistic operational class imbalance. Five supervised learning models—a Random Forest, Gradient Boosting, a Support Vector Machine, a Deep Neural Network, and a stacked ensemble—were benchmarked, with the ensemble embedded within the DT core representing the operational predictive model. Experimental results demonstrate strong performance, achieving an F1-score of 0.98 and an AUC of 0.995. The results confirm that the proposed DT–AI framework provides a scalable, interoperable, and cyber-resilient foundation for deployment-ready predictive maintenance in modern substation automation systems. Full article

Sugarcane (Saccharum spp.) is a globally important crop for sugar and bioenergy production. However, genetic improvement through conventional breeding is constrained by long breeding cycles, low genetic gain, and considerable operational complexity arising from its highly allopolyploid and aneuploid genome. With the increasing global demand for sustainable food and renewable energy, sugarcane breeding programs must accelerate the development of high-yielding, stress-tolerant cultivars through the integration of advanced biotechnological tools with traditional breeding approaches. Recent advances in genetic engineering, genomic selection (GS), and high-throughput omics technologies, including genomics, transcriptomics, proteomics, metabolomics, and phenomics, have substantially enhanced the efficiency of trait improvement related to growth, development, yield, and stress resilience. The integration of multi-omics data enables the dissection of regulatory networks linking genotype to phenotype, improves predictive accuracy, and provides deeper insights into the molecular mechanisms underlying complex traits. These integrative approaches support more informed selection decisions and accelerate genetic gain in sugarcane breeding programs. This review synthesizes recent technological developments and their practical applications in sugarcane improvement. It highlights the strategic implementation of transgenic and genome-editing technologies, genomic selection, and multi-omics integration to enhance yield potential and resistance to biotic and abiotic stresses, thereby contributing to sustainable sugarcane production and global food and bioenergy security. Full article

Increasing global water scarcity has intensified the adoption of water reuse as a sustainable strategy, particularly in regions affected by drought and pressure on natural resources. This paper presents a systematic review of the application of Life Cycle Assessment (LCA) in water reuse projects, focusing on research trends, methodological approaches, and opportunities for improvement. A systematic search was conducted in Web of Science, ScienceDirect, and Google Scholar for studies published from 2020 onwards using combinations of the keywords “life cycle assessment”, “LCA”, “water reuse”, “water recycling”, and “wastewater recycling”. Twelve studies were selected from 57 records identified, based on predefined eligibility criteria requiring quantitative LCA of water reuse systems. The results reveal a predominance of European research, reflecting regulatory advances and strong academic engagement in this field. The most frequently assessed impact categories were global warming, eutrophication, human toxicity and ecotoxicity, highlighting the environmental relevance of reuse systems. Energy consumption and water transport were identified as critical hotspots, especially in scenarios involving long distances and fossil-based energy sources. Nevertheless, most studies demonstrate that water reuse is environmentally viable, particularly when renewable energy and optimized logistics are applied. The review also emphasizes the need to better integrate economic and social dimensions and to adapt LCA methodologies to local conditions. Overall, the findings confirm LCA as a robust decision-support tool for sustainable planning and management of water reuse systems. Full article

Renewable energy investment evaluation continues to rely predominantly on techno-economic and environmental criteria, while equity-related considerations remain weakly embedded within formal decision-support frameworks. Although recent research increasingly acknowledges social impacts, spatial constraints, policy uncertainty, and financing structures, these dimensions are rarely integrated in a systematic and operational manner into investment appraisal. This paper addresses this gap by advancing an equity-oriented conceptual framework for renewable energy investment evaluation. Using an integrative literature review combined with thematic analysis, the study synthesises insights from techno-economic assessment, multi-criteria decision-making, energy justice scholarship, and equity-focused modelling studies. The analysis demonstrates that existing evaluation approaches inadequately capture distributional impacts, accessibility constraints, differentiated vulnerability, and equity-adjusted risk. In response, the proposed framework systematises these equity dimensions and embeds them directly into the core logic of investment evaluation alongside conventional criteria. By consolidating fragmented research insights into a coherent evaluative structure, the study contributes to the literature by clarifying how equity can be operationalised within renewable energy investment decision-making. The framework provides a foundation for future empirical applications and supports more socially responsive and analytically robust investment evaluation. Full article

The maritime sector is undergoing a profound transformation driven by disruptive technologies and global decarbonization objectives, placing new demands on Maritime Education and Training (MET) systems. Equipping maritime professionals with competencies for low-carbon shipping is now as critical as technological advancement itself. This study examines how disruptive technologies can be effectively integrated into MET frameworks to support environmental sustainability, using Panama as a representative case study of a major flag and maritime service state. A mixed-methods approach was adopted, combining a structured literature review, expert surveys, and a multi-criteria decision-making analysis based on the Analytic Hierarchy Process (AHP). The findings reveal a significant misalignment between existing MET curricula and the competencies required for decarbonized maritime operations. Key gaps include limited training in alternative fuels, emissions measurement and reporting, energy-efficient technologies, digital analytics, and regulatory compliance. Stakeholders also reported fragmented training provision, uneven access to emerging technologies, and weak coordination between academia, industry, and regulators, particularly in developing contexts. The results highlight the urgent need for curriculum reform and stronger cross-sector collaboration to align MET with evolving technological and regulatory demands. The study provides an applied, evidence-based framework for MET reform, with insights transferable to other systems facing similar decarbonization challenges. Full article

In the last year, great advances in the treatment outcomes of advanced EGFR-mutant NSCLC have been achieved. Two combination regimens, amivantamab plus lazertinib and platin-based chemotherapy plus osimertinib, have yielded, in the phase III randomized trials named MARIPOSA and FLAURA 2, statistically and clinically significant improvements in overall survival compared with monotherapy with osimertinib. However, translation to clinical practice of these relevant results is challenging for two main reasons. The first is that we have no evidence-based tools to choose among the two combinations, except their different safety profiles. The second is that combinations are significantly more toxic than osimertinib alone. Thus, osimertinib remains an effective treatment with an excellent safety profile, perhaps to be considered as still the best option in the majority of elderly patients and in all patients that do not intend to trade-off an excess of toxicity with survival prolongment. The safety and efficacy characteristics of the three treatment options are the basis for a patient-tailored treatment choice, but in a significant proportion of patients, a personal and intimate approach to quality of life and survival prolongment is to be considered the main driver within a well-structured shared decision-making process. Full article

In the domain of algorithmic decision-making, non-Markovian dynamics manifest as a significant impediment, especially for paradigms such as Reinforcement Learning (RL), thereby exerting far-reaching consequences on the advancement and effectiveness of the associated systems. Nevertheless, the existing benchmarks are deficient in comprehensively assessing the capacity of decision algorithms to handle non-Markovian dynamics. To address this deficiency, we have devised a generalized methodology grounded in category theory. Notably, we established the category of Markov Decision Processes (MDP) and the category of non-Markovian Decision Processes (NMDP), and proved the equivalence relationship between them. This theoretical foundation provides a novel perspective for understanding and addressing non-Markovian dynamics. We further introduced non-Markovianity into decision-making problem settings via the History Aggregator for State (HAS). With HAS, we can precisely control the state dependency structure of decision-making problems in the time series. Our analysis demonstrates the effectiveness of our method in representing a broad range of non-Markovian dynamics. This approach facilitates a more rigorous and flexible evaluation of decision algorithms by testing them in problem settings where non-Markovian dynamics are explicitly constructed. Full article

Road infrastructure underpins Nigeria’s economic competitiveness, yet Public–Private Partnership (PPP) performance is constrained not by inadequate legislation but by persistent weaknesses in enforcement and governance. Transparency deficits across procurement, design management, certification, and toll-revenue reporting have produced chronic delays, cost overruns, and declining public trust. This study offers the first empirical investigation of blockchain–Building Information Modelling (BIM) integration as a transparency-enhancing mechanism within Nigeria’s PPP road sector, focusing on Lagos State. Using a qualitative design, ten semi-structured interviews with stakeholders across the PPP lifecycle were thematically analysed to diagnose systemic governance weaknesses and assess the contextual feasibility of digital innovations. Findings reveal entrenched opacity rooted in weak enforcement, discretionary decision-making, and informal communication practices—including biased bidder evaluations, undocumented design alterations, manipulated certifications, and toll-revenue inconsistencies. While respondents recognised BIM’s potential to centralise project information and blockchain’s capacity for immutable records and smart-contract automation, they consistently emphasised that technological benefits cannot be realised absent credible institutional foundations. The study advances an original theoretical contribution: the Codify–Condition–Capacitate framework, which explains the institutional preconditions under which digital governance tools can improve transparency. This framework argues that effectiveness depends on: codifying digital standards and legal recognition; conditioning enforcement mechanisms to reduce discretionary authority; and capacitating institutions through targeted training and phased pilots. The research generates significant practical implications for policymakers in Nigeria and comparable developing contexts seeking institution-aligned digital transformation. Methodological rigour was ensured through purposive sampling, thematic saturation assessment, and documented analytical trails. Full article

Background: The rapid ageing of the cardiovascular population has profoundly transformed clinical practice, with an increasing proportion of patients presenting advanced age, frailty, multimorbidity, and functional vulnerability. Conventional cardiology models, largely derived from younger and selected populations, often fail to adequately address the complexity of cardiovascular care in older adults. Despite the growing development of cardiogeriatrics, the core competencies required for contemporary geriatric cardiology practice remain insufficiently defined. Methods: This narrative review synthesises evidence from cardiology, geriatrics, heart failure, and the palliative care literature, complemented by clinical expertise in integrated cardiogeriatric care pathways, to identify key competencies relevant to the care of older adults with cardiovascular disease. Results: Four major domains of geriatric cardiology competencies were identified: (1) advanced cardiovascular expertise adapted to ageing physiology, frailty, and multimorbidity; (2) integration of comprehensive geriatric assessment into cardiovascular decision-making; (3) a dedicated cardiogeriatric communication mindset supporting shared decision-making under prognostic uncertainty; and (4) system-based competencies focused on multidisciplinary coordination, care transitions, and therapeutic proportionality. Conclusions: Defining the core competencies of the geriatric cardiologist is essential to addressing the clinical and organisational challenges of an ageing cardiovascular population. This framework provides a pragmatic foundation for clinical practice, education, and future research, supporting integrated cardiogeriatric care models aligned with patient-centred outcomes. Full article

Intelligent Transportation Systems (ITS) have progressively incorporated machine learning to optimize traffic efficiency, enhance safety, and improve real-time decision-making. However, the traditional centralized machine learning (ML) paradigm faces critical limitations regarding data privacy, scalability, and single-point vulnerabilities. This study explores FL as a decentralized alternative that preserves privacy by training local models without transferring raw data. Based on a systematic literature review encompassing 39 ITS-related studies, this work classifies applications according to their architectural detail—distinguishing systems from models—and identifies three families of federated learning (FL) frameworks: privacy-focused, integrable, and advanced infrastructure. Three representative frameworks—Federated Learning-based Gated Recurrent Unit (FedGRU), Digital Twin + Hierarchical Federated Learning (DT + HFL), and Transfer Learning with Convolutional Neural Networks (TFL-CNN)—were comparatively analyzed against a client–server baseline to assess their suitability for ITS adaptation. Our qualitative, architecture-level comparison suggests that DT + HFL and TFL-CNN, characterized by hierarchical aggregation and edge-level coordination, are conceptually better aligned with scalability and stability requirements in vehicular and traffic deployments than pure client–server baselines. FedGRU, while conceptually relevant as a meta-framework for coordinating multiple organizational models, is primarily intended as a complementary reference rather than as a standalone architecture for large-scale ITS deployment. Through application-level evaluations—including traffic prediction, accident detection, transport-mode identification, and driver profiling—this study demonstrates that FL can be effectively integrated into ITS with moderate architectural adjustments. This work does not introduce new experimental results; instead, it provides a qualitative, architecture-level comparison and adaptation guideline to support the migration of ITS applications toward federated learning. Overall, the results establish a solid methodological foundation for migrating centralized ITS architectures toward federated, privacy-preserving intelligence, in alignment with the evolution of edge and 6G infrastructures. Full article

Background: Non-small cell lung cancer (NSCLC), particularly in advanced stages, has poor prognosis. The main objective of the study is to evaluate real-world treatment outcomes in advanced NSCLC patients harboring an EGFR mutation and being treated with TKIs. Methods: The EGFR mutation status was ascertained by next-generation sequencing. The observational cohort study used prospectively maintained registry data. Patient data were collected at two high-volume institutions in Austria between November 2020 and February 2025. The prevalence of EGFR mutations was 11% (145 out of 1267 patients). Results: Among 53 patients (stage IIIB or higher) with an EGFR mutation, median overall survival (OS) and median progression-free survival (PFS) were 17.7 months (95% CI: 10.4–24.9) and 14.2 months (95% CI: 7.4–20.9), respectively. A total of 36 patients harbored common EGFR mutations (exon 19 deletion or L858R point mutation) and exhibited a significantly better OS than those with an uncommon EGFR genotype (p < 0.005). Patients with exon 19 deletion (n = 25) showed the longest mOS, followed by those with L858R mutation (32.5 vs. 17 months). In multivariable analysis, the EGFR common mutation subtype (HR = 3.71 95%CI: 1.23–11.2) was associated with better OS. Patients with common EGFR genotypes, especially exon 19 deletion obtained longer OS and PFS compared with those with uncommon mutations in exon 18–21. Conclusions: The results underscore the prognostic role of distinct EGFR genotypes and the urgency of determining the mutation status in non-small cell lung cancer patients to ensure the best treatment decision. The study also highlights the challenges regarding to EGFR uncommon mutations and the resulting need for further research to investigate alternative treatment options. Full article

Background: Accurate preoperative staging is the cornerstone of therapeutic decision-making in gastric cancer (GC), yet standard modalities often fail to capture the full extent of disease, particularly in diffuse and poorly cohesive histotypes. This review aims to provide a comprehensive update on diagnostic imaging for GC, evaluating the established roles of CT, EUS, and PET/CT alongside the emerging capabilities of Magnetic Resonance Imaging (MRI) and Artificial Intelligence (AI). Methods: A structured narrative review was conducted by searching indexed biomedical databases for studies published between 2015 and 2024. A structured literature search screening process identified 410 relevant studies focusing on T, N, and M staging accuracy, quantitative imaging biomarkers, and radiomics. Results: While Multidetector CT remains the universal first-line modality, its sensitivity declines in infiltrative tumors and low-volume peritoneal carcinomatosis. EUS retains superiority for early (T1-T2) lesions but may offer limited value in advanced stages. Conversely, MRI (leveraging diffusion-weighted imaging (DWI) and multiparametric protocols) indicates superior soft-tissue contrast, potentially outperforming CT in the assessment of serosal invasion, nodal involvement, and occult peritoneal metastases. Furthermore, emerging fibroblast activation protein inhibitor (FAPI) PET tracers show promise in overcoming the limitations of FDG in mucinous and diffuse GC. Finally, radiomics and deep learning models are providing novel quantitative biomarkers for non-invasive risk stratification. Conclusions: Contemporary GC staging requires a tailored, multimodality approach. Evidence supports the increasing integration of MRI and quantitative imaging into clinical workflows to overcome the limitations of conventional techniques and support precision oncology. Full article