21684 research outputs found
Sort by
Investigating bifunctional protein degraders to silence PDE4 activity in disease
No full textAbstract not currently available
Suicide risk in sexual and gender minority adults: understanding the role of protective factors and minority stress
Get PDFAbstract available at each chapter
Rapid sequencing-based detection of bacterial species and their antimicrobial resistance genes for improved patient outcomes in bloodstream infections
No full textAbstract not currently available
Reframing intercultural education in edu-business: A decolonising approach
Get PDFThe ever-evolving global context has brought an increased focus on nonformal learning organisations such as edu-business which provides learning services as a ‘for-profit’ enterprise (Ball, 2012). However, edu-business is shaped by the neoliberal environment in which it is embedded. Its underlying economic drivers comprise the core challenges for learning developers in this sector. Moreover, there is a remarkable absence of research specific to edu-business which has raised questions on its capability in a globally diverse learning context. This study aimed to close that research gap slightly by concentrating on the specific learning area of intercultural education and the possibilities of a decolonising approach to curriculum and pedagogy. Specifically, the focus was on how knowledge is validated, how it is disseminated and by whom in edu-business. The overall study aim was not to provide a decolonising solution or set of guidelines for edu-business but to initiate a dialogue on decolonising intercultural education for edu-business. This study interrogated the pedagogical approaches to intercultural education for edu-businesses in Europe by exploring issues related to ‘quality of content’ (Krishna, 2009) for curriculum and pedagogy.
Using a qualitative research design, the data was collected from semi-structured interviews with learning developers in European-based edu-businesses. A framework analysis drew from concepts delineated by Shahjahan, Estera, Surla and Edwards’ (2022) decolonising curriculum and pedagogy (DCP) framework for higher education which provided a relevant comparison for exploring a decolonising approach to intercultural curriculum and pedagogy in edu-business. The research findings highlighted issues related to the contextual challenges, interrelated interpretations and the actualisation of decolonising pedagogy which have implications for how learning developers recognise conditions and relations of power in which intercultural learning is embedded. It requires their capacity to acknowledge their complicity, subjectivity and pedagogical agency in the propagation of dominant Eurocentric approaches which sanction power relations and have an impact on the quality of intercultural programmes in edu-business. The learning developers’ unique positionality has the potential to transform intercultural knowledge production into a more critically relevant and humancentric pedagogical practice as part of the greater decolonising project
Methodological developments in data fusion for lake water reflectance from satellite sensors
Get PDFFusing satellite-sensed reflectance data from different sources is of interest to monitor lake water quality, and the satellite sensors have possibly different spatial, temporal and spectral supports. The nonparametric statistical downscaling (NSD) model is an existing state-of-the-art fusion model which can account for a change of spatial and temporal support between two remote sensors [Wilkie et al., 2019]. However, the NSD model is computationally demanding for large datasets and does not allow multivariate responses with an additional spectral dimension. Thus, the aim of this thesis is to improve the computational efficiency of the NSD model and then extend this model to provide an approach that is suitable for a multivariate response to enable the fuse of reflectance data with different spectral and temporal supports from two sensors. The NSD model assumes that the discrete data at each location within a lake from each data source are observations of smooth functions over time and that the coefficients of these smooth functions are modelled as spatially correlated via a covariance matrix. In this thesis, a novel approach proposes using a Gaussian predictive process to approximate the spatial varying coefficients in the NSD model, which requires the inversion of a matrix with smaller dimensions in the Gibbs sampling process and hence reduces the computational time for the parameter estimation. The predictive performance and computational efficiency of the proposed nonparametric statistical downscaling model with Gaussian predictive process (NSD-GPP) are compared to the NSD model through simulation and using satellite reflectance data from Lake Garda. It was found that the NSD-GPP model achieves a similar predictive performance as the NSD model using less computational time. To enable data fusion from the two sensors with a multivariate wavelength dimension, a novel method using the two-dimensional B-spline basis functions was developed where the basis functions were used to represent the reflectance over both time and wavelength at each location, and a different precision parameter was used for each wavelength. Lake Garda is used as an example of interest here, and methods are general for any lake of interest in principle. Overall, it is found that the proposed multivariate NSD-GPP model could be used to make predictions for the unobserved wavelengths and time points within the observed range. It may be beneficial to provide reflectance data at higher temporal and wavelength frequencies, and this model could in principle be extended to consider similar challenges in space
Numerical study of morphing helicopter rotors
Get PDFThis thesis shows high-fidelity numerical simulations for combined fluid, structural and servo-systems of a helicopter, to make predictions and validations of helicopter rotor on blade-actuators. This aero-servo-elastic approach forms the main novelty of this thesis. The physics of the active rotor blade aerodynamics and structural dynamics are simulated, and the resulting knowledge can then be implemented in comprehensive rotor suites. Assessing the physics is important for the introduction of new rotor concepts, something which low- and mid-order methods can not do. The Helicopter Multi-Block 3 (HMB3) simulation suite developed at University of Glasgow was expanded to include the servo system in the aeroelastic models. The software does not include a rigid body model for the rotor hub. Both hub and blade structural dynamics are simulated in the commercial FEM solver MSC NASTRAN.
Expanding the flight envelope with higher top-speeds, while avoiding retreating blade stall, has been a challenge since the inception of helicopters. A promising innovation is the variable twist rotor blade, which through intrinsic actuation, can optimise the twist for the given flight condition and exploit aeroelasticity to modify the rotor blade path for vibration control and possibly higher trim-able thrust. This thesis only considers twist actuators, due to their benefits in weight, simplicity and airworthiness over other actuation methods. However, the aero-servo-elastic modelling can be applied to any rotor blades with piezoceramic actuators, such as active flaps or Gurney flaps, if the actuator is sufficiently resolved in the structural model.
The aeroelastic method is validated against the Helicopter Validation and Acoustic Baseline rotor (HVAB). Performance metrics, structural deformations and flow physics are compared and confirm the accuracy of the method. Additionally, rotor blade structural models of 1D beam and 2D/3D-finite elements are modelled and compared, including a mesh study. Finite elements allow the modelling of on-blade actuators, instead of simplifying the problem to only include their effects in simulations. The structural models of the Smart Twisting Active Rotor (STAR) are compared to experimentally obtained data, for validation of the methods used in this thesis. This also includes a study of the actuator modelling via a thermal analogy method, where the voltage on a piezoelectric material is equated to temperature in a thermally expanding material.
Using the developed aero-servo-elastic method, improvements in the forward flight vibration metric could be predicted for the STAR rotor blade in two flight conditions. The largest vibration improvement was observed at high-speed level flight. In this condition, increased passive blade twist has a large vibration penalty. In a level flight at maximum thrust, experiencing blade-vortex interactions (BVI), the active twist was shown to improve the moment trim, leading to conclusions of higher thrust capacity in this relatively high-speed flight case. The observed effects on the rotor lift-to-drag ratios were negligible.
For the STAR, a 1D beam model and a mixed 2D/3D finite element model were compared. The beam model overpredicts the blade mode frequencies of the measured rotor blades, because the sectional properties of the received dataset were slightly overpredicted. A modified set of properties, yielding more accurate mode frequencies is also presented. The finite element model, built from available geometric and material data, was within 10% of the measured frequencies. The strong 3-dimensional structural coupling of the finite element model is showcased, which could not be observed in the beam model. A small blade-untwisting under centrifugal force is found for the finite element model. The actuator effectiveness degrates under centrifugal tension and this mechanism is discussed. A hovering simulation of the rotor, with a loosely coupled fluid-structure simulation shows the differences between the models. It concludes, that especially for blades with torsion actuators, 3D finite element model approaches are necessary to obtain correct results
How effective is China’s SO2 emissions regulation?
Get PDFChina’s rapid industrialization has spurred severe environmental challenges, with sulfur dioxide (SO₂) emissions posing significant threats to public health and sustainable development. This thesis evaluates the effectiveness of China’s SO₂ emissions regulations, focusing on the 2006 Eleventh Five-Year Plan (FYP) policy as a quasi-natural experiment. Leveraging difference-in differences (DID) methods and comprehensive firm-level data, the study examines the policy’s causal impacts on firm pollution emissions, total factor productivity (TFP), two-way foreign direct investment (FDI), and export performance.
The analysis reveals that the SO₂ emissions regulation achieved measurable success in reducing firm SO2 emissions, particularly among state-owned and large-scale firms, though with heterogeneous effects across regions and industries. Strikingly, while stricter environmental mandates initially raised compliance costs, they stimulated total factor productivity gains by incentivizing technological innovation. The policy also reshaped China firms’ global engagement: stricter regulation reduced inward FDI in but encouraged outward FDI (OFDI) as firms sought cleaner technologies abroad. Meanwhile, export performance exhibited nuanced outcomes. Export value and export product quality improved as firms upgraded production processes to meet environmental standards.
Mechanism tests underscore the role of innovation offsets, reduction cost, and market reallocation in driving these outcomes. Heterogeneity analyses highlight divergent responses based on firm ownership, firm size, and industry intensity, offering insights into the uneven distribution of regulatory costs and benefits. By integrating environmental economics with firm-level dynamics, this research contributes empirical evidence on the trade-offs and synergies between environmental governance and economic performance in emerging economies
Enhancing data representation in distributed machine learning
Get PDFDistributed computing devices, ranging from smartphones to edge micro-servers—collectively referred to as clients—are capable of gathering and storing diverse types of data, such as images and voice recordings. This wide array of data sources has the potential to significantly enhance the accuracy and robustness of Deep Learning (DL) models across a variety of tasks. However, this data is intrinsically heterogeneous, due to the differences in users’ preferences, lifestyles, locations, and other factors. Consequently, it necessitates comprehensive preprocessing (e.g., labeling, filtering, relevance assessment, balancing, etc.) to ensure its suitability for the development of effective and reliable models. Therefore, this thesis explores the feasibility of conducting predictive analytics and model inference on edge computing (EC) systems when access to data is limited, and on clients’ devices through federated learning (FL) when direct access to data is entirely restricted.
The first part of this thesis focuses on reducing the data transmission rate between clients and EC servers by employing techniques such as data and task caching, identifying data overlaps, and evaluating task popularity. While this strategy can significantly minimize data offloading to the lowest possible level, it does not entirely eliminate dependence on third-party entities.
The second part of this thesis eliminates the dependency on third-party entities by implementing FL, where direct access to raw data is not possible. In this context, node and data selection are guided by predictions and model performance. The objective is to identify the most suitable nodes and relevant data for training by clustering nodes based on data characteristics and analyzing the overlap between query boundaries and cluster boundaries.
The third part of this thesis introduces a mechanism designed to support classification tasks, such as image classification. These tasks present significant challenges when building models on distributed data, particularly due to issues like label shifting or missing labels across clients. To address these challenges, the proposed method mitigates the impact of imbalances across clients by employing multiple cluster-based meta-models, each tailored to specific label distributions.
The fourth part of this thesis introduces a two-phase federated self-learning framework, termed 2PFL, which addresses the challenges of extreme data scarcity and skewness when training classifiers over distributed labeled and unlabeled data. 2PFL demonstrates the capability to achieve high-performance models, even when trained with only 10% to 20% labeled data compared to the available unlabeled data.
The conclusion chapter underscores the importance of adaptable learning mechanisms that can respond to the continuous changes in clients’ data volume, requirements, formats, and protection regulations. By incorporating the EC layer, we can alleviate concerns related to data privacy, reduce the volume of data needing offloading, expedite task execution, and facilitate the training of complex models.
For scenarios demanding stricter privacy-preserving measures, FL offers a viable solution, enabling multiple clients to collaboratively train models while adhering to user privacy protection, data security, and government regulations. However, due to the indirect access to data inherent in FL, several challenges must be addressed to ensure the development of high-performance models. These challenges include imbalanced data distribution across clients, partially labeled data, and fully unlabeled data, all of which are explored and demonstrated through experimental evaluations
Quantitative analysis of the collective movement and migratory behaviour of Atlantic salmon
Get PDFThe migration of Atlantic salmon (Salmo salar ) is a complex ecological process of great importance to conservation and ecosystem management. Despite a rapid decline in the Atlantic salmon population, little is known about the fine-scale behaviour of juvenile migration to the sea, which occurs in an environment under strong anthropogenic pressure and results in high mortality.
Juvenile salmon migration involves large quantities of small animals travelling underwater through a complex riverine landscape. It is difficult to study, but recent advances in tracking technology and mathematical methods now allow the behaviour and ecology of downstream migrating salmon to be explored in depth.
This thesis reviews recent advances in the study of Atlantic salmon, and develops further the mathematical and computational methods of movement ecology. These help in forming hypotheses about how the fish behave, leading to laboratory experiments that focus on their responses to flow conditions, social influences during obstacle navigation, and the behavioural differences between wild and hatchery-reared individuals
Impounded waters pose a challenge for migrating smolts due to the lack of strong flow to provide directional cues. In laboratory experiments, I established the baseline flow values that prompt a behavioural response in salmon smolts, necessary for the design of river structures. In another experiment, I showed evidence of collective decision-making in navigating obstacles during movement in an experimental flume. This finding emphasises the density-dependent factors in migration success and necessitates further study of the collective behaviour of this species that have thus far been mostly under-explored. I also show clear differences in behaviour of hatchery animals compared to wild ones, providing guidance on the usability of hatchery smolts in further studies and design of river infrastructure.
Modern machine learning methods allow improved analysis of data in many contexts. Improvements to visual tracking of fish based on deep-learning models are presented, allowing for detailed analysis of movement in laboratory and field experiments where video cameras are becoming ever more prevalent. In this thesis, I present a new visual tracking method that is tailored to correctly predict the movement of animals and tested on simulated data inspired by common movement models.
The Bayesian modelling framework of Approximate Bayesian Computation is leveraged to analyse movement patterns from acoustic telemetry data. This simulation-based method combines the hypotheses about the fine-scale movement with computational methods that allow efficient parallelisation on GPU-accelerated hardware.
This thesis provides insights into the migratory behaviour of Atlantic salmon smolts, with significant implications for conservation efforts, ecological engineering applications, and the design of effective river infrastructure. The findings emphasise the necessity of considering social behaviours and the differences between wild and hatchery fish in both modelling and practical implementations to aid in the preservation of this important species
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇