26572 research outputs found
Sort by
Deep Probabilistic Surrogate Modelling for Uncertainty Quantification in Mangrove Hydro-morphodynamics
Full text linkMangrove ecosystems are increasingly recognised as essential nature-based solutions (NbS) for enhancing coastal resilience against sea level rise and climate-induced extreme events. However, achieving robust uncertainty quantification (UQ) for hydro-morphodynamic models of mangrove systems remains a critical and unresolved challenge. The inherent complexity of physical processes, coupled with the computational demands of solving Navier–Stokes partial differential equations (PDEs), complicates conventional UQ approaches. Traditional surrogate models, such as Gaussian Processes (GPs), often fall short in capturing the non-Gaussian behaviour and high-dimensional interactions characteristic of coastal dynamics, while physics-informed neural networks (PINNs), though promising, face scalability issues that limit their application in large-scale uncertainty quantification (UQ). To overcome these limitations, we introduce an efficient and scalable probabilistic framework based on Deep Gaussian Processes (Deep GPs), which hierarchically stack multiple GP layers to capture complex, multi-scale, and non-Gaussian dependencies that conventional surrogate models fail to represent. The proposed Deep GP model significantly reduces computational cost by over three orders of magnitude (≈ 5 × 10³ times faster; ≈ 1.4 min vs > 5 days for the full numerical solver) while maintaining high predictive accuracy (fivefold improvement; RMSE = 0.0095 m vs 0.0465 m for standard GP), enabling reliable propagation of uncertainty across complex, nonlinear system dynamics. Through application to a high-resolution mangrove model, we demonstrate the framework’s potential to support evidence-based planning for climate adaptation and ecosystem-based coastal resilience. This work offers a novel pathway to integrate advanced UQ into operational decision-making for sustainable coastal management
The African Medicines Agency: historical perspective of its origins, evolution, institutional structure and future prospects
Full text linkBackground: The African continent has long faced fragmented regulatory systems, resulting in delayed access to safe, effective, and quality-assured medical products. To address these challenges, the African Medicines Regulatory Harmonisation (AMRH) Programme was launched in 2009 by the African Union, laying the groundwork for the establishment of the African Medicines Agency (AMA). The AMA represents one of the most significant continental developments to harmonize regulatory practices, improve access to quality-assured medical products, and strengthen public health systems across Africa. Objectives: The objectives of this review were to examine the historical development of AMA, its Treaty and proposed institutional framework, as well as operational pilots such as the Continental Listing of Human Medicinal Products implemented by the AMRH since 2023. Methods: A narrative literature review approach was used, sourcing official African Union documents, peer-reviewed publications, and technical reports from African Union Commission, AUDA-NEPAD, WHO, and AMRH stakeholders published between 2005 and 2025. Results: The AMA was formally established by treaty adopted by the AU heads of states and governments in 2019 and entered into force in November 2021. As of June 2025, 31 AU Member States had ratified the Treaty. The agency’s governance and organizational structure include a Conference of State Parties, Governing Board, Secretariat, and Technical Committees. Pilot projects such as the AMRH Continental Listing demonstrated the feasibility of reliance mechanisms, though challenges remain in national legal harmonization, funding, and capacity disparities. Conclusion: The AMA represents a transformative step toward regulatory convergence in Africa. While challenges persist, the Treaty framework and pilot outcomes provide a strong foundation for its operationalisation and the long-term success in improving medical product regulation and public health across the continent
Fuzzy Superpixel Segmentation with Anisotropic Total Variation Regularization
Full text linkThis paper presents a superpixel segmentation algorithm that integrates anisotropic total variation regularization within a fuzzy clustering framework. While isotropic total variation is well-known for its edge-preserving properties, its non-adaptive nature often leads to over-regularization. In contrast, the anisotropic model formulates superpixel regularity in relation to image contours, thereby preventing the loss of image details in areas of high contour density during optimization. Compared to classical segmentation algorithms that employ non-adaptive regularization, the proposed content-adaptive approach enhances superpixel regularity while maintaining boundary adherence to image contours. Furthermore, to optimize the functional effectively, an alternating direction method of multipliers along with the enhanced Chambolle’s fast duality projection algorithm are employed. Competitive experiments against existing regular segmentation algorithms demonstrate that our proposed methodology achieves superior performance in terms of boundary recall, compactness, and shape regularity criteria, outperforming these methods by an average of at least 3%, 5%, and 3%, respectively. Furthermore, when compared with irregular segmentation algorithms, our approach achieves the best results in terms of compactness, contour density, and shape regularity criteria, with average improvements of at least 56%, 22%, and 45%, respectively
SpectR*: Expander Graph Based Seller Recommender for Mitigating Cold Start Loss in Larger Facebook Market
Full text linkThe enormous growth of the Facebook marketplace is prompting sellers to leverage the network to enhance their business. The pandemic has introduced more cold start sellers on Facebook, and the buyers' pandemic needs have made imperfect observations on seller Item quality. Users' trust in the sellers varies from quality to quality and from point to point, and thereby could lead to a loss from the cold start sellers. The need of the hour is an approach which could walk along the paths of strongly connected graphs to analyse products brought by the users, and thereby, better recommendations can be built. This article presents a new graph based method, the SpectR*, to recommend a seller to other Facebook business groups. Due to the demographic structural differences, larger networks have smaller expansion capabilities, which is why many recommender systems failed on graph networks like Facebook and Twitter, to analyse the reviews from the users. Our proposed work is based on constructing an Expander Graph on Facebook user network, based on calculated Cheegers constant. We used this Graph theory based Cheegers constant as measure for deriving well connected Facebook family of Graphs on which the product reviews can be analysed
Fencing and Wheelchair Fencing Research: A Scoping Review Dataset
No full textA scoping review of all experimental studies published on fencing and wheelchair fencing. The data set has all the details of the papers from authors, to year of publication and key findings
Mixed-methods assessment of engagement with a digital intervention: the Wrapped feasibility Randomised Controlled Trial
Full text linkDigital health behaviour change interventions can face the challenge of low participant engagement, which can limit intervention effectiveness. Mixed methods approaches to understanding engagement, which capture the online and offline behaviours of participants, as well as the cognitive and affective aspects of engagement, are infrequently reported. The aim of this study was to explore these aspects of engagement for a digital intervention (Wrapped) to enable its optimisation ahead of testing in a randomised controlled trial (RCT). Wrapped is a digital intervention that aims to increase correct and consistent condom use, thereby decreasing the incidence of sexually transmitted infections among young people aged 16-24 years. Analytics data and website user history data were combined with data from surveys and qualitative interviews. Together this data was examined to assess the behavioural, cognitive, and affective aspects of engagement with the intervention. Results showed that participants experienced few barriers during the registration process, but that the tailoring questions used to assign content to individual users may not have been working as intended. Pre-determined intervention goals were as follows: ordering Sample Pack 48 (60.8%), using Condom Ordering Service 37 (52.9%), ordering Condom Carrier 31 (49.2%), watching Condom Demo video 7 (10%), watching Discussing Condoms video 4 (6.8%), and watching Real Life video 9 (13.6%). Participants described their use and enjoyment of the products they ordered; notably the condom carrier was less well liked and used. Participants reported not engaging with the video components, either because they were unaware that they existed or because they expected to find watching them to feel awkward. This study demonstrates that taking a mixed-methods approach to studying engagement provides a more complete understanding of where and how digital interventions need to be optimised than using single methods in isolation; this in turn is likely to lead to more effective interventions. Trial registration: ISRCTN Registry ISRCTN17478654
Autapses enable temporal pattern recognition in spiking neural networks
Full text linkMost sensory stimuli are temporal in structure. How action potentials encode the information incoming from sensory stimuli remains one of the central research questions in neuroscience. Precise spike timing is known to represent information in spiking neuronal networks, yet the information processing mechanisms of spiking neuronal networks is poorly understood. One feasible way to understand the processing mechanism of a spiking network is to associate the structural connectivity of the network with the corresponding functional behaviour. This work demonstrates the structure-function mapping of spiking networks evolved (or handcrafted) for a temporal pattern recognition task. The task is to recognise a specific order of the input signals so that the output neuron of the network spikes only for the correct placement and remains silent for all others. The minimal networks obtained for this task revealed two complementary roles of autapses in recognition. First, autapses enable a seamless transition to the next network state when a new input signal arrives. Second, in the absence of the input signal, they allow the network to maintain a network state for an extended period, a form of memory. To show that the recognition task is accomplished by transitions between network states, we map the network states of a functional spiking neural network (SNN) onto the states of a finite-state transducer (FST). Finally, based on our understanding, we define rules for constructing the topology of a network handcrafted for recognising a subsequence of signals in a particular order. The analysis of minimal networks recognising patterns of different lengths revealed a positive correlation between the pattern length and the number of autaptic connections in the network. Furthermore, in agreement with the behaviour of neurons in the network, we were able to associate specific functional roles of ’locking,’ ’switching,’ and ’accepting’ to neurons
An Online Pilates Program for People with Hypermobility: A Pragmatic Clinical Trial Looking at Function, Interoception, Kinesiophobia, and Physical Activity Levels
Full text linkPurpose: This study evaluated whether an independent, online modified Pilates program alters function, interoception (internal body awareness), activity levels, and kinesiophobia (fear of movement) in people with symptomatic joint hypermobility. Patients and Methods: This pragmatic clinical trial included and exercise group that did 8 weeks of modified Pilates and an 8-week waitlist control group. People with symptomatic hypermobility were asked to do an independent, online Pilates program designed specifically for people with hypermobility; each module was about 25 minutes, and participants were asked to do at least 3 days/week. Outcome measures included the Bristol Impact of Hypermobility (BIoH), International Physical Activity Questionnaire (IPAQ), Revised Body Awareness Questionnaire (BARQ), and Tampa Scale of Kinesiophobia (TSK). Clinical Trial # NCT07118865. Results: A total of 420 participants completed questionnaires at 8 weeks: 200 completed 8 weeks of Pilates and 220 were on a waitlist. The Pilates group demonstrated statistically significant improvements in BIoH, BARQ, and TSK immediately after the intervention (p< 0.001 for each) and compared to the control group (p< 0.001 for each). Improvements in the Pilates group remained statistically significant at 6 months. IPAQ did not change for any group. Conclusion: This online Pilates program for people with symptomatic hypermobility improved BIoH, BARQ, and TSK, though changes were modest. The exercises did not improve IPAQ. Online exercise may provide a cost-effective way to encourage life-long activity in people with hypermobility. Limitations include the inability to monitor performance of the Pilates, high drop-out rates, and the inability to control for changes in other treatments patients may have received
Numerical investigation on cooling performance of phase change assisted direct ventilation system for data center
No full textTraditional data centers often use mechanical cooling systems, leading to high energy consumption and waste of natural cooling resources. Thus, a novel phase change ventilation device that combines natural cooling with phase change storage has been designed to maintain continuous natural cooling of the data center by storing cold energy from the natural cold air using phase change plates (PCPs), and eliminate the reliance on mechanical refrigeration in traditional data centers and achieving energy savings. In this study, the cooling performance of the proposed device is numerically analyzed and the feasibility of the model is verified by experiments, filling the research gap in data centers for this method. Considering main effectors, i.e., the inlet air velocity (IAV), inlet air temperature (IAT), phase change plate thickness (PCPT), phase change temperature (PCT), and thermal conductivity of encapsulation material (TCEM) on the cooling performance of the device. The results show that: (1) Phase change ventilation device can reduce the IAT of 34 °C by an average of 2.53 °C within 8 h (2) When the IAV increases from 1 m/s to 4 m/s, the average cooling performance of the phase change ventilation device decreases by 62.93 %. (3) In the phase change latent heat stage, the temperature difference (TD) of phase change ventilation device decreases almost linearly over time. (4) The significance analysis of orthogonal experiment shows the impact of various factors on the cooling performance of phase change ventilation device as follows: IAV > IAT > PCPT > PCT > TCEM
Women’s intimate partner violence victimization and healthcare access barriers in Nigeria: the moderating role of education
Full text linkBackground: This study examined how women’s educational attainment influences healthcare access in the context of intimate partner violence (IPV) in Nigeria. It explored whether IPV and education independently predict healthcare access barriers and whether education moderates the IPV–barriers relationship, thereby clarifying education’s role in autonomy and healthcare access. Methods: Data were drawn from the 2018 Nigeria Demographic and Health Survey and included a weighted sample of 7,553 women aged 15–49 years. Four perceived barriers of needing permission, obtaining money for treatment, distance to facilities, and going alone were assessed alongside measures of IPV and education. Weighted logistic regression models tested the main and moderating effects of IPV and education, adjusting for relevant sociodemographic covariates. Results: The respondents’ mean age was approximately 33 with a standard deviation of 8.23. Women experiencing IPV were significantly more likely to report barriers related to permission (AOR = 1.37, p < .001), financial constraints (AOR = 1.76, p < .001), and distance (AOR = 1.26, p < .001). Secondary (AOR = 0.54, p < .01) and higher education (AOR = 0.13, p < .01) attenuated the effect of IPV on permission barriers, while the same educational levels (AOR = 0.75, p = .042) reduced IPV-related financial barriers. Conclusions: IPV substantially heightens barriers to women’s healthcare access, whereas education serves a protective role, particularly for autonomy-related barriers such as permission and financial constraints. As the first national-level analysis in Nigeria to demonstrate education’s moderating role in the IPV–healthcare link, this study highlights the need for integrated health and education policies that empower women, promote IPV screening, and address structural and cultural barriers to care