78146 research outputs found
Sort by
Would 'secure' users lead to secure commons? Surprisingly not! : A framework to evaluate effective power and collective outcomes in cybersecurity
No full textIndividuals are often held responsible when adverse cyber incidents occur. The ensuing narrative, explaining the occurrence, points to confounding factuals e.g., inability to act securely, a convincingly deceptive attack, or selfishness/laziness. The underlying assumption is that: if only humans were different (acted securely), such adverse events would not occur. In this paper, we use a game theoretic approach to investigate the counterfactual in cyber: what would happen if individuals were indeed different? To that end, we propose a generic framework drawing upon two games. Our proposed framework can help move the field towards judicious responsibilization of individuals and eliminate knee-jerk scapegoating. We use this framework to examine a specific social harm - data pollution. Our explorations show that even if individuals always behaved securely, this would not necessarily improve collective outcomes. We show that individuals are sometimes not in a position to change security outcomes, however secure their behaviours. The proposed framework can be applied to highlight those entities that are indeed in a position to influence security outcomes in the wider aggregate harm landscape. Future research should build on our work to assign responsibilities in the cyber domain, explore ways to operationalise the games to carry out empirical research, and contribute to novel paradigms such as ethical responsibilization in the context of data breaches
Multiphysics analysis of a flexible oscillating water column wave energy converter with dielectric elastomer membrane
Get PDFFlexible wave energy converters (FlexWECs) have emerged as a promising solution to address the limitations of conventional rigid devices in harsh marine environments. Among them, oscillating water column (OWC) systems integrated with dielectric elastomer generators (DEGs) offer simplified architectures, enhanced adaptability, and direct wave-to-electric energy conversion. However, the complex multiphysics interactions between fluid, structure, and electric fields remain poorly understood, hindering design optimization and performance prediction. This study develops a high-fidelity computational framework to simulate the coupled fluid-structure-electric behaviour of a flexible OWC wave energy converter (WEC) with a DEG membrane. The framework is first validated against experimental data, demonstrating good agreement in capturing the deformation of the flexible membrane induced by the coupled electrostatic and hydrodynamic forces. Subsequently, the model is applied to investigate how electric field influences the WEC system behaviour under regular wave excitation. Results show that applying an electric field reduces the effective stiffness of the membrane, leading to increased deformation. Additionally, it does raise overall structural stress levels, especially near the membrane centre and edge regions, where the maximum stresses are observed. Notably, electric excitation induces a secondary deformation mode in the membrane during the near-flat phase. These effects become more pronounced with increasing initial voltage, which also leads to an approximately quadratic increase in output power. The insights gained from this study provide a deeper understanding of fluid-structure-electricity (FSE) interactions in flexible OWC WECs and offer design guidance for enhancing energy harvesting efficiency in next-generation WEC devices
Comparing microscopic and macroscopic diffusion in drug delivery : a study of small drug and protein dynamics in a supramolecular peptide hydrogel
Get PDFOver recent decades, medicine development has increasingly turned to biologics for their higher efficacy and reduced side effects compared to small molecule drugs. However, the necessity for parenteral administration of these labile compounds poses significant challenges, requiring frequent injections and highlighting the need for sustained release systems to improve patient adherence. Drug delivery vehicles are currently regarded as acting as a diffusion barrier and, particularly supramolecular gels, offer a promising solution to this challenge due to their tunable properties and biocompatibility. This study focuses on diffusion within a peptide supramolecular hydrogel based on the ultra-short Fmoc-diphenylalanine (FmocFF) moiety, exploring its potential as a drug delivery carrier. By examining dynamics at different timescales, we aim to decouple steric and non-steric effects of the fibre network on solute diffusion. Using Quasi-Elastic Neutron Scattering (QENS), we investigate solvent and gel network dynamics, and the picosecond self-diffusion behavior of various drugs to focus solely on hydrodynamic interactions. We further assess bulk diffusion over 12 hours by in vitro release studies using the Subcutaneous Injection Site Simulator (SCISSOR), mimicking possible interactions occurring during drug release. Our results demonstrate that from molecular-level to bulk diffusion, hydrodynamic interactions are mitigated and masked by factors such as steric confinement and surface erosion. This work highlights the necessity for more systematic data to bridge short-time to long-time scale bulk diffusion mechanisms to rationally tune drug release and ultimately design better drug delivery vehicles
A palladium inverse crown : synthesis and characterisation
Get PDFA unique palladium inverse crown complex in [Cs2(18-crown-6)2]2+ [(PdPtBuPh)6(PPh)]2− is presented where a neutral (PdP)6 ring hosts the dianionic PhP2− guest. Characterised by SC-XRD and high-resolution mass spectrometry, its solution constitution in THF is probed by DOSY and multinuclear NMR with quantitative insight into its bonding by DFT calculations
Optimising neural network hyperparameters by predicting unseen architectures with learning curve approach
No full textHyperparameter optimisation (HPO) is a critical yet computationally expensive task in training neural networks. While recent research has used learning curve prediction to terminate poor configurations early, no existing method can predict the full set of unseen learning curves within a search space. Current techniques often rely on meta-learning across datasets or partial curve information, limiting both generalisability and precision. This study introduces SEquential LEarning Curve Training (SELECT), a novel HPO approach that accurately predicts the full learning curves of all unseen hyperparameter configurations using a sequence prediction model trained on a subset of the same dataset. SELECT employs a new representation of learning curve data—converted into structured sequences with aligned starting points—and leverages a Convolutional Gated Recurrent Neural Network (CGRNN) for high-fidelity forecasting. Benchmarked across multiple real-world regression datasets, SELECT consistently outperformed established methods including Random Search (RS), Hyperband (HB), Gaussian Process Bayesian Optimisation (GPBO), and Tree-structured Parzen Estimator (TPE), achieving up to 20% improvements in predictive accuracy while maintaining consistent and predictable computation time. Its design enables full parallel training of configurations, making it highly suitable for large-scale or time-sensitive applications. In addition to its performance and scalability, SELECT offers a unique advantage in search space visualisation. By predicting entire learning curves, it allows practitioners to inspect the shape and structure of the optimisation landscape, revealing patterns such as diminishing returns and performance plateaus. This transparency enhances trust, interpretability, and decision-making in both research and industrial HPO workflows
Ruling with ideology : politicians' beliefs and privatizations
Get PDFI examine how political ideology shapes China’s privatization wave around 2000. Using a novel ideology measure constructed based on belief formation mechanisms, I show that provincial governors with above-median communist beliefs privatize 2.8% fewer state-owned enterprises (SOEs) each year than their below-median-belief colleagues. Firms privatized under such governors also achieve fewer efficiency gains after the sale. Mechanism analysis finds evidence that this performance gap arises from ideologically driven choices: pro-communist governors are more likely to privatize weaker or less important firms, adopt transaction structures associated with inferior outcomes, and manage subsidies in ways that exacerbate post-privatization challenges. Moreover, ideology moderates how governors learn from experience and respond to signals, shaping their adaptation in managing privatization. Jointly, these results demonstrate that political leaders’ beliefs can substantially influence the trajectory and outcome of major economic reforms. It also highlights the importance of ideology in driving politicians’ decisions under authoritarian regimes
How useful are outcrop samples to constrain subsurface thermal properties for geothermal exploration? Case study from the Chester formation, UK
Get PDFHot sedimentary aquifers (HSAs) offer a sustainable option for geothermal energy, but exploration risks remain a major challenge. Drilling is capital intensive, especially for geothermal projects, and early in project appraisal it is often uncertain whether a formation will provide sufficient thermal performance to justify investment. Outcrop samples have therefore gained attention as analogues for subsurface reservoir properties, providing low-cost preliminary data before committing to drilling. While this approach is well established in the hydrocarbon industry, its application to geothermal settings has been limited. Existing geothermal studies have largely focused on structural, petrophysical, or mineralogical characterisation, while systematic comparisons of these properties and thermal properties between outcrop and core samples remain rare. This is a critical gap, as thermal properties govern the long-term performance of a geothermal system. This study compares seventeen outcrop and twenty-one core samples (taken from depths of 22–96 m) from the Triassic Chester Formation in the Cheshire Basin, UK, to assess the transferability of outcrop-derived data. The samples were analysed for mineralogy, petrophysical, and thermal properties. Results show that while outcrop samples exhibit some variability due to weathering at the Earth’s surface, they generally have similar mineralogical and petrophysical properties to core samples. Thermal conductivity is higher in cores (2.15 ± 0.01 W m⁻¹ K⁻¹) compared to outcrops (1.74 ± 0.01 W m⁻¹ K⁻¹). Variations are also observed between different lithotypes, with pebble-rich samples generally exhibiting higher thermal conductivity than massive, layered, or cross-bedded sandstones. These findings highlight the potential of outcrop samples for early-stage geothermal exploration. They provide a cost-effective alternative to drilling for constraining rock properties, particularly thermal conductivity, at depth, while offering insights into depositional environment and lithology. This integrated approach could improve subsurface predictions, reduce exploration costs, and support the growth of the geothermal sector by de-risking investment decisions
Prescribing for older people with sensory impairment : a qualitative interview study with independent prescribers in primary care
Get PDFObjectives To explore prescribers’ awareness of medicine-related challenges of older people (≥65 years) with sensory impairment (hearing, visual or dual impaired) and identify the influences on prescribing behaviours for these patient populations. Design Semistructured interviews were completed online. Setting Primary care-based prescribers in the UK. Participants Independent prescribers working in primary care. Participants were recruited through professional networks and organisations, social media and using snowballing. Purposive sampling was used to ensure variation in roles, practice/organisational settings and geographical location. Results 15 prescribers participated, including general practitioners (n=6), pharmacists (n=5), nurses (n=3) and one optometrist. Many demonstrated limited awareness of sensory impairment and suggested that outdated patient records contribute to it being easy to overlook. Prescribers underestimated sensory impairment prevalence, with one predicting that only a small proportion of older patients had hearing loss. Formal training on prescribing for older people with sensory impairment was minimal, and most relied on experiential learning. Prescribers employed strategies to support safe prescribing, such as simplifying regimens and selecting lower-risk medications. The prescribers also reported a lack of evidence-based guidelines or resources tailored to these patient populations. Conclusions Prescribers currently receive minimal training to support their prescribing practices for older people with sensory impairment. Given the increasing prevalence of age-related sensory impairment, evidence-based resources and training are needed to support prescribing for these patient populations
Particle engineering of a needle-like active pharmaceutical ingredient into size controlled agglomerates : Part II. Evaluating direct compression manufacturability, microstructure and product performance
Get PDFSpherical agglomeration presents a promising particle engineering strategy for converting crystalline active pharmaceutical ingredients (APIs) with challenging bulk powder properties into manufacturable drug products. However, a critical gap remains in understanding their downstream processability via direct compression and the resulting drug release performance. This study provides a comprehensive assessment of how a successfully spherical agglomerated API, prepared at controlled particle sizes, influences tablet mechanical behaviour, microstructure arrangement, and dissolution performance. Three agglomerate size fractions (35 µm, 88 µm, and 143 µm), together with the original un-agglomerated API, were each formulated at a fixed 20% w/w drug load using an identical direct compression excipient blend (FastFlo 316, Avicel PH-101, AcDiSol, Ligamed MF-2 V) and evaluated alongside a benchmark wet granulated formulation. Compactability and compressibility modelling demonstrated systematic trends in tablet strength and densification across all formulations, with agglomerate size exerting only a minor influence on compactability. Advanced Raman imaging and nano-CT, integrated with data-driven tablet microstructure analysis, revealed that smaller agglomerates and un-agglomerated particles promoted more homogeneous API distributions, whereas larger agglomerates formed more localized clusters within the tablet matrix. Post-compression in-tablet particle and pore size distributions showed agglomerate domains were largely preserved during compaction. In-vitro dissolution studies further showed that all spherical agglomerate formulations achieved rapid drug release, meeting the label requirement of Q = 75% within 45 min, with a modest particle size dependence observed under biorelevant dissolution conditions across gastric, duodenal, and jejunal compartments. Overall, the results demonstrate that under the tested particle sizes, spherical agglomerates were suitable for direct compression solid dosage form manufacturing
The motherhood penalties : insights from women in UK academia
Get PDFThe motherhood penalty is often understood as a salary differential between mothers and non-mothers. We use an original survey of academic women in the UK to understand whether the motherhood penalty extends to other dimensions of a woman's career and experience in the workplace. We explore these penalties via an original survey of academic women in the UK. Becoming a mother, we show, has no effect on salary, but slows down career progression. Mothers report higher levels of job satisfaction yet indicate heightened perceptions of gendered salary unfairness. We then explore several factors potentially mitigating the motherhood penalties. On the formal side, more generous maternity provisions are associated with higher salaries, and longer childcare hours facilitate career progression. On the informal side, a sympathetic Head of Department boosts job satisfaction. At home, having a supportive partner plays a key role in mothers' professional success. Our paper highlights the varied penalties mothers encounter even in a highly skilled profession, and the necessity of a multi-faceted policy response
Access Repository Dashboard
Do you manage University of Strathclyde Institutional Repository? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!