13506 research outputs found
Sort by
Improved Deep Neural Networks for Generative Robotic Grasping
Get PDFThis thesis provides a thorough evaluation of current state-of-the-art robotic grasping methods and contributes to a subset of data-driven grasp estimation approaches, termed generative models. These models aim to directly generate grasp region proposals from a given image without the need for a separate analysis and ranking step, which can be computationally expensive. This approach allows for fully end-to-end training of a model and quick closed-loop operation of a robot arm.
A number of limitations are identified within these generative models, which are identified and addressed. Contributions are proposed that directly target each stage of the training pipeline that help to form accurate grasp proposals and generalise better to unseen objects. Firstly, inspired by theories of object manipulation within the mammalian visual system, the use of multi-task learning in existing generative architectures is evaluated. This aims to improve the performance of grasping algorithms when presented with impoverished colour (RGB) data by training models to perform simultaneous tasks such as object categorisation, saliency detection, and depth reconstruction. Secondly, a novel loss function is introduced which improves overall performance by rewarding the network to focus only on learning grasps at suitable positions. This reduces overall training times and results in better performance on fewer training examples. The last contribution analyses the problems with the most common metric used for evaluating and comparing offline performance between different grasping models and algorithms. To this end, a Gaussian method of representing ground-truth labelled grasps is put forward, which optimal grasp
locations tested in a simulated grasping environment.
The combination of these novel additions to generative models results in improved grasp success, accuracy, and performance on common benchmark datasets compared to previous approaches. Furthermore, the efficacy of these contributions is also tested when transferred to a physical robotic arm, demonstrating the ability to effectively grasp previously unseen 3D printed objects of varying complexity and difficulty without the need for domain adaptation. Finally, the future directions are discussed for generative convolutional models within the overall field of robotic grasping
There and back again - mobility and burial rites in early medieval England c. AD 400-1000: an analysis of strontium isotopes in cremated human remains
Get PDFObservable changes in funerary rites in early medieval Britain culminate in the re-appearance of cremation in the 5th and 6th centuries, and again briefly in the 9th century. These changes are traditionally linked to questions of mobility and encompass many thousands of individuals. Analyses making use of isotopic systems have been immensely successful in the exploration of inhumed, unburnt remains in the last few decades - touching on the diet, mobility, and the lifecycle of individuals, and recent work on cremated remains has shown that some of these isotope analyses are also reliable when conducted on calcined bone. Following on from this work, this thesis aims to test whether strontium isotope analysis on cremated remains can be used to extract new information from early medieval cemeteries to investigate such aspects as mobility, animal-human relationships, and social practice of the cremating communities. The project uses multi-skeletal sampling to explore the mobility history of cremated individuals buried at the 5th to 7th century mixed-rite cemeteries at Ingleby Barwick, Stockton-on-Tees, and Cleatham, North Lincolnshire, as well as at the 9th century Scandinavian barrow cemetery at Heath Wood, Ingleby, Derbyshire. Plants within 25km catchments around the sites were sampled to establish the variation of bioavailable 87Sr/86Sr, ameliorating the sample density and therefore precision of existing Sr isoscapes.
The application of strontium isotope analysis uncovered what appears to be regional mobility amongst the cremating communities at Cleatham and Ingleby Barwick, and long-distance mobility from Scandinavia at Heath Wood. At Cleatham, the larger sample of individuals (54 humans and 4 animals) also allowed us to observe that women, born outside the 25km catchment, moved more frequently into the area than men. This was noted to pertain especially to the early phase of use of the cemetery and highlights that trends in mobility may change across time. Animal-human relationships were explored with reference to the strontium isotope results and agency/personhood theory at all three case-study sites. This work is the first of its kind with a focus on the early medieval period in Britain and its results show that the method has enormous potential to unlock aspects of mobility, animal-human relationships, and social practice of the cremating communities
Essays on the economics of networks
Get PDFNetworks (collections of nodes or vertices and graphs capturing their linkages) are a common object of study across a range of fields includ- ing economics, statistics and computer science. Network analysis is often based around capturing the overall structure of the network by some reduced set of parameters. Canonically, this has focused on the notion of centrality. There are many measures of centrality, mostly based around statistical analysis of the linkages between nodes on the network. However, another common approach has been through the use of eigenfunction analysis of the centrality matrix. My the- sis focuses on eigencentrality as a property, paying particular focus to equilibrium behaviour when the network structure is fixed. This occurs when nodes are either passive, such as for web-searches or queueing models or when they represent active optimizing agents in network games. The major contribution of my thesis is in the applica- tion of relatively recent innovations in matrix derivatives to centrality measurements and equilibria within games that are function of those measurements. I present a series of new results on the stability of eigencentrality measures and provide some examples of applications to a number of real world examples
An investigation into recent outlet glacier dynamics within Vincennes Bay, Wilkes Land, East Antarctica.
Get PDFThe Antarctic Ice Sheet has been in a state of negative mass balance over recent decades, with mass loss largely occurring from the West Antarctic Ice Sheet. However, recent studies indicate that Wilkes Land, East Antarctica, has lost mass at accelerating rates over the past two decades, exhibiting a dynamic response to oceanic forcing. Overlying the marine-based Aurora Subglacial Basin (ASB), Wilkes Land has been referred to as the potential ‘weak underbelly’ of the East Antarctic Ice Sheet and is drained by several major outlet glaciers. Despite their potential importance, few of these glaciers have been studied in detail. This includes the six outlet glaciers draining into Vincennes Bay, a region recently reported to have the warmest intrusions of modified Circumpolar Deep Water (mCDW) ever recorded in East Antarctica. This thesis seeks to improve our understanding of the understudied Vincennes Bay outlet glaciers, providing a first overview of recent ice dynamics observed between 1963 and 2022. Optical satellite imagery, differential satellite synthetic aperture radar interferometry (DInSAR) techniques and a range of secondary datasets were employed in order to assess change across four fundamental glacier parameters: terminus position, ice surface velocity, ice surface elevation, and grounding line position. Decadal fluctuations between terminus advance and terminus retreat recorded across the Vincennes Bay outlet glaciers correspond closely with wider patterns reported across Wilkes Land, potentially correlated to variations in sea ice production. Ice surface velocities were generally stable between 2000 and 2021, with some fluctuations measured across the grounding line of Bond East Glacier. Observed changes in ice surface elevation were spatially variable, with a consistent, albeit relatively modest, thinning trend seen across Vanderford Glacier between 2003 and 2017. Enhanced rates of ice thinning were measured across each of the Vanderford, Adams, Anzac, and Underwood Glaciers between 2017 and 2020, potentially linked to the widespread decline in Antarctic sea ice extent reported during the austral spring of 2016. Most importantly, extensive grounding line retreat was observed at Vanderford Glacier, measured at 18.6 km between 1996 and 2020. Such rapid grounding line retreat (0.8 km yr-1) is consistent with the notion that warm mCDW is able to access deep cavities formed below the Vanderford Ice Shelf, driving high rates of basal melting. With an inland retrograde bed slope observed along the Vanderford Trench, such oceanic forcing may have significant implications for the future stability of Vanderford Glacier. This study shows that the dynamic response of Vanderford Glacier has been more muted than expected given the high magnitude of grounding line retreat observed. Enhanced thinning and the onset of ice flow acceleration may therefore be predicted over the coming decades
Optimizing a Rydberg Atom-based Terahertz Imaging System
Get PDFWe build upon previous work by characterising and optimizing the signal from a Rydberg atom-based terahertz (THz) imaging system, where atoms are excited to a Rydberg state by a three step ladder scheme. These Rydberg atoms provide a THz imaging mechanism via a fluorescence signal, . We implement a Pound-Drever-Hall lock and a digital scanning transfer cavity lock on the Rydberg laser. Both schemes see an improvement in the stability of fluorescence signal over time, to within 3\%. We characterise the effect of laser parameters on . A pumping scheme consisting of alternating circular polarisations for each step are found to maximise , providing a increase compared to the all-linear configuration. The dependence of on Rydberg laser detuning and THz detuning is also characterised, and splitting due to the AC Stark shift is observed at higher THz powers. Simultaneous resonant excitation of the Rydberg and THz transition is found to maximize . A repump laser is implemented on both the D1 and D2 lines, and found to increase most on the D1 line, by a factor of seven. The effect on of the repump lasers power and detuning is characterised, and a two-photon transition is observed in the D2 line spectrum. A change in spatial intensity distribution of the THz beam is observed under the effect of rotation of a linear polariser, which is theorised to be due to a vector diffraction pattern
Exploring student nurses' intercultural care experiences in clinical practice
Get PDFThis study explores how student nurses experience intercultural nursing care in their everyday clinical practice. It adopts a qualitative social constructionist approach to investigate their perceptions of and responses to positive experiences and the challenges they encountered in their nursing placements and to seek to understand intercultural competencies that are salient in facilitating effective intercultural communication.
Several key findings emerged from the study: 1) Participants socially construct their intercultural care experience in the six domains: intercultural encounters, knowledge, skills, attitudes, engagement and outcomes; and consider intercultural competence is a continual development process. 2) Even though the individual scope of intercultural care encounters (ICEs) varies, the participants all valued the opportunity to be exposed to ICEs, as these allowed them to negotiate and adjust to different cultures, values and worldviews. 3) The participants observed how cultural beliefs and behaviours affect individual patients’ interpretations or perceptions of their health condition, how they view its treatment and the efficacy of the medication. Without this essential knowledge, aside from the lack of understanding of a patient’s cultural needs, the treatment and care maybe inadequate or inappropriate, which can seriously affect patient’s outcomes. However, the data showed the gaps in this aspect of nursing education. 4) One of the biggest challenges the student nurses encountered in ICEs was linguistic barriers, since language affected not only the patients’ ability to express their needs and understand the healthcare instructions, but also had nursing implications. 5) Participants reported positive and challenging attitudes in ICEs, which affected their learning, wellbeing, and intercultural communication performance. When workplace adversity persists, some participants even considered leaving the course. It was through personal resilience, accessing support from academics and peers that they decided to stay in their chosen career. 6) The motivations to engage intercultural care were affected by internal and external factors. 7) The participants highlighted that without considering outcomes for the care recipient (patient) and care provider (the student nurse), competence in intercultural care cannot be validated.
An intercultural care competence ecological framework was developed, which provides a comprehensive view of intercultural care and highlights the dynamic nature of competence development. In these ways, this study complements existing cultural care theories.
The study offers valuable insights for researchers, educationalists, practitioners and policymakers in the areas of intercultural communication and interactions in healthcare services
Forward-looking information in Integrated Reporting: The influence of suppliers and customers
Get PDFThis thesis examines the influence of suppliers and customers on forward-looking information disclosure and the credibility of this disclosure practice under the Integrated Reporting (IR) context.
According to stakeholder theory and legitimacy theory, suppliers and customers hold critical resources for an organisation’s survival and growth; thus, organisations would consider suppliers’ and customers’ interests when making reporting decisions. Further, concerning signalling theory, organisations have incentives to signal both credible and misleading information. These theories underpin this study’s investigation throughout.
This thesis applies a multi-method approach which combines the method of content analysis and interview. To fulfil the research aims, this thesis utilises content analysis to examine a dataset of organisational reports published by 96 organisations from 5 countries. In addition, this thesis conducted 19 semi-structured interviews with personnel from organisations and stakeholders from supply chain consulting firms, accounting bodies, and standards-setting organisations.
The findings reveal that organisations incorporate suppliers’ and customers’ interests into forward-looking information disclosure to discharge accountability, address legitimacy exposures, and secure resources for survival. However, the extent to which organisations address suppliers’ and customers’ information demands is related to the management’s awareness of supplier and customer inclusiveness, regulations, and reporting framework. Furthermore, suppliers and customers influence an organisation’s business and strategy developments, and these developments are reflected in the forward-looking information disclosure. To enhance the credibility of forward-looking information concerning suppliers and customers, organisations have implemented multiple mechanisms (e.g. data measurement and data monitoring) throughout the reporting process. However, this disclosure practice lacks completeness and rigorous materiality determination processes. Finally, IR has an impact on the adoption of the approach to including suppliers and customers when making forward-looking information disclosure and the credibility of this disclosure practice.
This thesis brings together three streams of research on forward-looking information disclosure (motivations, inclusivity, and credibility) in a single study. The findings of this thesis confirm the importance of developing a supplier and customer inclusive approach to reporting. Yet, this reporting approach has not been fully adopted, and relevant reports are not credible enough for stakeholders to rely on, although IR plays some role in enhancing this reporting practice. This thesis argues that the current practice of disclosing forward-looking information concerning suppliers and customers is at risk of being a “box-ticking” activity, rather than being thoroughly understood and implemented as a reform of reporting approaches. From the investigation of both organisations’ and stakeholders’ perspectives on this disclosure practice, this thesis offers management, standard-setters, and academics novel insights into the role of suppliers and customers in corporate reporting which has received little academic attention
Bayesian Approaches to Emulation for a Complex Computer Crop Yield Simulator with Mixed Inputs
Get PDFAgriculture is one area where the simulation of crop growth, nutrition, soil condition and pollution could be invaluable in any land management decisions. The Environmental Policy Integrated Climate Model (EPIC) is a simulation model to investigate the behaviour of crop yield in response to changes in inputs such as fertiliser levels, soil, steepness, and other environmental covariates. We build a model for crop yield around a non-linear Mitscherlich Baule growth model to make inferences about crop yield response to changes in continuous input and factor variables. A Bayesian hierarchical approach to the modelling was taken for mixed inputs, requiring Markov Chain Monte Carlo simulations to obtain samples from the posterior distributions, to validate and illustrate the results, and to carry out model selection.
The emulation of complex computer simulations has become an effective tool in exploring this high-dimensional simulated process's behaviour. Initially, we built a Bayes linear emulator to efficiently emulate crop yield as a function of the simulator's continuous inputs only. We explore emulator diagnostics and present the results from the emulation of a subset of the simulated EPIC data output. Computer models with quantitative inputs are used widely, but the challenge is incorporating the factors. We propose a framework for solving this issue considering the Bayes linear emulation approach. We explore a variety of correlation structures to represent the mixed inputs and combine this with the Bayes linear approach to construct an emulator. Finally, we developed a method to make an optimal decision for the farmers to gain maximum utility considering yield and pollutants, accounting for weather factors, land characteristics and fertiliser use
Electricity Sector Reforms in Sub-Saharan Africa
Get PDFElectricity sector reforms has been one of the most transformative energy sector policies in sub-Saharan Africa (SSA) in modern history. However, studies on reforms in SSA remain limited, with quantitative analysis almost non-existent. This thesis contributes to the literature on electricity sector reforms in SSA through an assessment of reform performance and its connection to key electricity sector topics in the region including investments and productive efficiency, access, and cost efficiency. The thesis is structured in a three-paper format, with each paper focused on each of the key challenges mentioned.
In Paper One, I assess the performance of electricity sector reforms in 37 SSA countries between 2000 and 2017 using a parametric multi-input multi-output distance function and a Stochastic Frontier Analysis (SFA) approach. From this assessment, I found an effective reform model in SSA to involve vertical unbundling with an electricity law, a sector regulator, and private ownership and management of electricity assets where desirable. I also found reforms to be positively correlated with efficient investments in generation but negatively correlated with reduction in technical network losses. On the institutional front, perceptions about non-violent institutional features such as corruption and governance effectiveness were found to have no significant relationship with reform performance whereas negative perceptions about terrorism and violence were found to be negatively correlated with reform performance.
In the Second Paper, I examined the determinants of electricity access performance in 46 SSA countries from the viewpoint of reforms using a production function and SFA from 2000 to 2017. I found generation capacity adequacy and the efficiency with which electricity is produced and used to be positively correlated with the rate of access expansion. The wealth of countries was also found to be positively correlated with improved electrification outcomes while the wealth of households in a country was found to be negatively correlated with inefficiencies that interfere with electrification efforts. The reform step that was found to engender these positive electrification outcomes was the presence of a sector regulator, while unbundling and private sector participation were found to be negatively correlated with access performance.
In the Third Paper, I explored the relationship between reforms and cost of electricity services to provide an economic perspective to issues of cost under-recovery in SSA electricity systems. Through a synthesis of reform theories and case studies and using small electricity systems as a surrogate for liberalised electricity systems without competitive markets, I showed the connection between reforms and costs. I made a case for a structural approach to issues of cost under-recoverability in SSA electricity systems leveraging contestability opportunities in mobile powerplants in generation, yardstick competition in distribution and retail and regional integration of electricity markets
Systemic Risk Based Portfolio Selection
Get PDFThis dissertation examines portfolio selection under systemic risk using performance measures. In the first chapter, we propose a novel performance measure to construct optimal portfolios that explicitly incorporate the occurrence of systemic event. Investors maximize an ex-ante modified Sharpe ratio that is conditional on some systemic event, with the latter interpreted as a low market return environment. We solve the portfolio optimization problem analytically under the absence of short-selling constraint and numerically when short-selling constraint is imposed. The approach is made operational by embedding it in a multivariate dynamic setting via dynamic conditional correlation and copula models. In the second chapter, we further enhance the portfolio selection approach proposed in the first chapter by using machine learning techniques. Specifically, the optimal portfolio is solved through a three-step supervised learning model. First, the smooth pinball neural network is employed to predict conditional marginal return distribution. Secondly, we use copula to model dependence between portfolio assets and the market, based on which we generate return scenarios. Lastly, we maximize the ex-ante conditional Sharpe ratio based on simulated returns. Unlike the previous chapter, where we use statistical models to forecast return distributions, in this chapter we take advantage of a distributional machine learning model along with a set of predictors that includes more than 1,000 predictive signals. In the last chapter, following the similar idea of conditional Sharpe ratio, we propose another systemic risk-based performance measure namely the conditional Rachev ratio. This measure inherits the advantage of unconditional Rachev ratio in the sense that it can account for asymmetric information of portfolio return distribution. Moreover, we build a link between our new measure and the well-know CoVaR measure in the finance literature. In each chapter, we construct a comparative analysis using data on the US stock market. Overall speaking, all the backtesting results demonstrate the superiority of our proposed approaches against popular benchmark strategies in terms of profitability and systemic risk, where the outperformance is robust to the inclusion of transaction costs
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! 👇