7440 research outputs found
Sort by
Life at the front: defining cyanobacterial traits for plasticity in changeable ocean
No full textTheoretical thesis.Bibliography: pages 50-54.Chapter 1: Introduction -- Chapter 2. Experimental design, materials and method -- Chapter 3: Results -- Chapter 4. Discussion .Microbial prokaryotes and eukaryotes thrive in the oceans where they represent the majority of biomass. They have adapted to and exploit almost every available niche by employing a wide diversity of functional traits. It is well known that distinct genetic lineages of marine picocyanobacteria occupy large geographical areas, such as the vast subtropical ocean gyres, or defined latitudinal slices encompassing the temperate, mesotrophic regions. However, a big proportion of the ocean corresponds to regions that are dynamic, such as the neritic coastal zone, or environments influenced by boundary currents. The aim of this thesis is to identify whether there are lineages of picocyanoacteria that are specifically adapted to variable ocean regions using comparative genomics and ecology. Secondly, to identify what specific molecular mechanisms they utilise to adapt. And finally, build and refine an efficient molecular toolbox to genetically manipulate the lineages of picocyanobacteria that could, for example, experimentally define the role of putative genes involved in adaptations to dynamic ocean regions . In the light of rapidly changing environmental conditions, a molecular - ecological understanding of how these abundant model cyanobacteria adapt will provide important insight s into the future productivity and the health of the ocean in general.1 online resource (vi, 64 pages) colour illustration
Structural investigation of learning communities in higher education online learning
No full textTheoretical thesis.Bibliography: pages 147-163.Chapter 1. Introduction -- Chapter 2. Learning communities in higher education -- Chapter 3. Social network analysis and learning communities in higher education online learning: a systematic literature review (Paper One) -- Chapter 4. Investigating virtual communities of practice with social network analysis: guidelines from a systematic review of research (Paper Two) -- Chapter 5. Social network analysis: a framework for identifying communities in higher education online learning (Paper Three) -- Chapter 6. Influence of learning design on the formation of online communities of learning (Paper Four) -- Chapter 7. Identifying online communities of inquiry in higher education online learning (Paper Five) -- Chapter 8. Conclusion.The unprecedented growth in online learning over the past decade has led to similar unprecedented interest in online learning communities. Given their long - established pedagogical significance, learning communities in one form or another have featured prominently in the field of education. Two types have communities that have gained considerable popularity in higher education online learning (HEOL) include the communities of practice (CoPs) and communities of inquiry (CoIs) which originate from the Community of Practice and Community of Inquiry frameworks. There is no shortage of literature reporting on the myriad benefits of learning within CoPs and CoIs which are commonly used as guides for designing learning environments and as theoretical lenses for assessing pedagogical processes in the traditional, face - to - face and online learning environments. This research presents a novel approach for investigating the structure of online communities and identifying CoPs and CoIs in HEOL using social network analysis (SNA).The CoP and CoI frameworks are contextually and conceptually distinct, the former developed specifically for online learning research and practice and the latter originating in the context of professional learning of apprentices. However, both types of communities comprise of structural and non - structural components. The structural component constitutes the interactions that take place between individuals within the community. In HEOL, these interactions typically occur within a learning management system (LMS). These interactions represent the paths or channels upon which the non - structural or qualitative components of the frameworks emerge. The proliferation of online learning has brought with it access to large amounts of data which is increasingly being used by researchers for investigating the design, cultivation, and sustainability of CoPs and CoIs in HEOL. However, a large majority of the existing research on CoPs and CoIs relies heavily on extensive and time - consuming qualitative methodologies and therefore focuses on the non-structural components of the frameworks . To date, although there is some literature that explores certain structural aspects of CoPs and CoIs, there have been no attempts at conceptualization of the holistic structure of and unique structural differences between CoPs and CoIs. Therefore, much remains to be understood with regards to the structural aspects and implications thereof of the two types of communities - a glaring gap in existing literature. Not only that, by virtue of being retrospective, qualitative research does not allow for realization of the immediate practical value of the CoP and CoI frameworks, a short-coming identified in existing literature.Thus, pivoting on the assumption that the distinct conceptual underpinnings and somewhat different pedagogical orientations of the CoP and CoI frameworks would be reflected in the interactional or structural dynamics within the communities, this research explores if and how SNA can be applied to online data for structurally investigating and identifying CoPs and CoIs in HEOL. This thesis comprises of a total of eight chapters, five of which are stand - alone research papers, each based on a different set of data. The thesis follows a sequential mixed method approach inclusive of two exploratory systematic literature reviews, a methodological framework and multiple case studies ...1 online resource (170 pages) colour illustration
Privacy preserving distributed data processing and game theoretic methods for data utility and privacy tradeoffs
No full textEmpirical thesis.Bibliography: pages 55-62.1. Introduction -- 2. Literature review -- 3. Privacy preserving federated learning -- 4. Game theory in privacy preserving machine learning -- 5. Conclusions and future research -- Appendix -- Bibliography.In this thesis, we focus on differential privacy and the trade-off between privacy level and accuracy of shared databases. For differential privacy, the Laplace mechanism is commonly utilised, for which the choice of value for a privacy parameter ε, also known as the privacy budget, plays a key role in the trade-off between privacy level and accuracy. We aim to build a game-theoretical model of choosing this privacy budget ε, while optimising the utility of the shared databases to which the differential privacy mechanism is applied. In this thesis, we consider differentially private queries applied in the context of two different models. The first model is an information-theory based query system using a discrete Laplace mechanism.The utility and leakage are quantified by information theory with min-entropy between original information, real answer and reported answer in the system. The second model is based on data analysis with privacy-aware machine learning using differentially-private gradient queries. We quantify the quality of the trained model by fitness cost, which is a function of differential-privacy parameters and the size of the distributed datasets, to capture the trade-off between privacy and utility by machine learning. Then, game theory is used to analyse the utility-leakage tradeoffs for both of these two models respectivelyMode of access: World wide web1 online resource (x, 62 pages) diagrams, graph
Towards cooperatively enhanced resonant dipole forces in optical levitation, harnessing quantum emitters in diamonds
No full textEmpirical thesis.Bibliography: pages 201-218.1. Introduction -- 2. Optical trapping -- 3. Investigating the dipole force on nitrogen vacancy centres in nano-diamonds -- 4. Photo-dynamics of nitrogen vacancy centres in nano-diamonds -- 5. Silicon vacancy centres in nano-diamonds -- 6. Conclusion and future work -- Appendix -- List of symbols -- References.Optical trapping offers a non-contact, non-destructive tool for manipulating and handling particles, from micron sized particles down to individual atoms with light. One can distinguish two different applications for optical trapping. The first relying on the ability to trap small particles (from tens of nm to tens of µm); the second, related to the confinement and cooling of atoms or collections of atoms. Until now there has been no system that combines the forces due to both dielectric particle trapping and atom trapping as they are usually contained in completely separate parameter regimes, even though both of these trapping applications arise from the same force. The force is the result of the interaction of the polarisability of the trapped object with the electric field of the trapping laser, which depends directly on the dipole strength trapped object. In the case of classical trapping, the force acts on an induced dipole caused by the electric field on the object, whereas in the case of atom trapping, the induced dipole arises from the optical transitions present in the electronic structure.In this way, optically trapped nano-particles containing embedded optical defects brings about a new regime of trapping and cooling nano-particles. This new trapping regime will allow well established techniques from atom trapping to be applied to the more massive nano-particles. The result is an optical trap showing capabilities with improved trapping strengths and cooling mechanisms. The enhanced control will allow us to cool the centre of mass motion of nano-particles down to their quantum ground state, even in a room temperature environment, exciting for high precision sensing and macroscopic quantum experiments.In this thesis I focus primarily on identifying and observing the trapping forces on the optical defects inside of optically trapped nano-diamonds. I developed an experimental set-up and procedure to isolate, trap and reliably measure the trapping behaviour of nano-particles as a platform for investigating the atomic forces on optical defects in nano-diamonds.I continue with a detailed analysis of the photo-physics of the Nitrogen Vacancy centre which is the most interesting and studied optical defect in diamond. I explore the implications of the intense trapping laser field, required for optical trapping, on the feasibility of observing defect related forces on NV centres in nano-diamonds. Due to the limitations of the Nitrogen Vacancy defect I explore the properties of Silicon Vacancy centres in diamond, highlighting the properties that an ideal particle would posses, with a particular emphasis on generating collective effects. I investigate the effects of the high intensity, near infra-red trapping laser field and show the exciting possibilities of cooling the centre of mass motion and the internal temperature of the nano-particle simultaneously.Whilst I have not yet reached the goal of levitating a particle and observing the cooperatively enhanced resonant optical dipole forces on its internal embedded defects, I have undertaken a number of steps providing the ground work and foundations for ultimately observing and using these forces.Mode of access: World wide web1 online resource (xxii, 218 pages) colour illustration
Millimeter wave beam alignment using deterministic compressive sensing
No full textEmpirical thesis.Bibliography: pages 50-54.1. Introduction -- 2. An overview of compressive sensing -- 3. System model -- 4. Pilot beam pattern design -- 5. Simulation results -- 6. Conclusion -- References.Designing the beamforming vectors for channel estimation in mmWave systems is challenging becauseof the narrow beams required and the small number of useful directions. The state of the art employs random or structured random beamforming to leverage compressive sensing techniques to solve this problem using a small number of measurements. In this dissertation, inspired by existing deterministic sensing matrices from the theory of compressive sensing, two novel deterministic compressive sensing approaches are proposed for channel estimation in mmWave systems. In the proposed approaches, the Kronecker product or row-by-row Kronecker product of existing deterministic sensing matrices are used to design the structure of pilot beam patterns for the beam alignment process. These approaches not only result in significant overhead reduction, but also present improvement in terms of performance for some scenarios.Mode of access: World wide web1 online resource (viii, 54 pages) colour illustration
Functional graphs of polynomials and their twist over finite fields
No full textEmpirical thesis.Bibliography: pages 43-44.1. Introduction -- 2. Building experimentation tools -- 3. Connected graph analysis -- 4. Hamiltonian cycle analysis -- 5. Discussion -- 6. Conclusion.Recently, functional graphs, i.e., graphs generated from a polynomial function over Finite Fields Fp, have received renewed interest, due to their applications in Computer Science. In this thesis, we study new functional graphs of degree two generated by a polynomial and its twist.We perform a computational analyses of these functional graphs by developing new algorithms, and optimizing their implementation performances (including multi-threading). Our results show that (i) most graphs are strongly connected or have only two components (including one giant component and one small component of two or three vertices); (ii) every connected component of the graph have many Hamiltonian cycles (growing exponentially with the finite field); (iii) these Hamiltonian cycles can be used to construct balancing binary sequences. Also the Hamiltonian property makes these graphs distinct to well-known random mappings where the expected cycle length is about √p.These experimental results were used to guide several theoretical analysis and then compared with the relevant mathematical proofs.Mode of access: World wide web1 online resource (44 pages) graphs, table
Modelling the impact of lifeline infrastructure failure during natural hazard events
No full textTheoretical thesis.Bibliography: pages 159-175.Chapter 1. Introduction -- Chapter 2. Compounding impacts of lifeline failure and the inclusion of lifeline disruption in current disaster plans -- Chapter 3. Applying graph theory to assess lifeline network disruption during natural hazard events: an example using the Tokyo Subway -- Chapter 4. Exposure of roads to volcanic ash from a future eruption from Mount Fuji, Japan: implications for evacuation and clean-up -- Chapter 5. Discussion -- References -- Appendices.In the case of natural hazard-caused disasters, direct impacts including building damage and loss of life are relatively well studied. Indirect disruption, on the other hand, including supply chain disruption and business interruption is harder to predict, quantify, visualise and insure. The need to better prepare for indirect disruption comes from the increasing cost and interruption it causes. A component of indirect disruption is the failure of lifelines such as power, communication, transportation and water; critical infrastructure and essential services that modern society relies on for everyday living. The disruption of lifeline services during natural hazard events has the potential to impact populations by exacerbating the hazard itself and/or hindering the ability to respond to or recover from the event. Lifeline failure can also propagate outside the reach of the hazard footprint, causing disruption in regions not directly impacted by the event. In preparation for the true impacts of natural hazard events on society there is a need to better understand the exposure of lifeline infrastructure, the interconnectedness and behaviour of lifeline networks and to identify vulnerable populations that rely on their operation. Current research on lifeline networks focus efforts on the evaluation of network characteristics, their optimisation and robustness to random failure, or the consequences of targeted attacks. Limited research has been undertaken on the impacts of natural hazard events on these systems and the flow-on effects of failure for disaster response and recovery. This thesis utilises mathematical graph theory tools alongside natural hazard modelling to analyse and quantify the extent of lifeline disruption during natural hazard events and the flow on effects of service failure. A future eruption of Mount Fuji in Japan is used as the major case study scenario to assess the usefulness of graph theory techniques in aiding disaster mitigation, emergency response and community recovery. In particular graph theory was used to assess the impacts of ash fall on the evacuation plans for Yamanashi Prefecture with regards to a future 1707 Hoei type eruption. It was found that: Ash induced road closures have the potential to affect current evacuation plans for Yamanashi Prefecture, particularly for those residents who are set to evacuate at or after the onset of a future eruption. Ash fall accumulation on roads, even after a few hours from the onset of an eruption, can inhibit road use, resulting in long detours or the inability for residents to be able to evacuate unassisted. After the cessation of an eruption, ash fall can impact the return of evacuees to their homes by either blocking roads or damaging buildings, affecting safety. Evacuees will have to wait for roads to be cleared of ash, and buildings to be assessed for damage, before they are able to return. In an eruption scenario where wind conditions are predominantly westerly the current plan for residents to evacuate to the north east of Yamanashi prefecture is not advisable. Assigned host locations in the northeast would be impacted by ash fall themselves; adding additional pressure on these communities and potentially resulting in further evacuations. This scenario provided the opportunity to test graph theory techniques in natural hazard risk assessment and to demonstrate how graph theory can assist post event recovery in a real world context. Methods developed in this study can be used to further explore impacts of ash fall, or other volcanic phenomena, in other prefectures around Mount Fuji or other volcanoes throughout Japan. Moreover, these methods can be used to address the exposure and risk to lifelines from other natural hazard events or even to compare between them. The results of this thesis show that graph theory techniques, alongside Geographic Information Systems tools and hazard modelling, with an understanding of the use and vulnerability of particular lifelines, can help to envisage potential problems that could result from lifeline failure and aid in the process of recovery. Not only is it important to make lifeline infrastructure more resilient to disruption from future natural hazard shocks, there is also a need to increase resilience by preparing communities to cope with service outages. For true shared responsibility to occur, local governments and communities need to be better informed and prepared so they can cope with the absence of lifelines during a disaster. Collaboration between all stakeholders is required to bridge information gaps and to create holistic disaster scenarios in order to provide more realistic and accurate assessments of future natural hazard impacts -- abstract.1 online resource (xxii, 210 pages
Low temperature confocal spectroscopy in bright nanodiamonds
No full textEmpirical thesis.Bibliography: pages 59-64.I. Introduction -- 2. Colour centres in diamond -- 3. Confocal build -- 4. Characterising nanodiamond samples -- 5. Conclusion -- Appendix -- References.While colour centres in diamond have numerous exciting room-temperature applications, it remains important to be able to study and control their properties at cryogenic temperatures where crystal phonon activity is suppressed. A cryogenic confocal microscope has been developed using galvanometer beam scanning and a closed-cycle helium cold-finger cryostat. This has required novel software control, and a new set of confocal scanning modules have been developed for the Qudi experimental control software suite. Testing has shown remarkable galvanometer repeatability and stability over many hours. The high scanning speed and low thermal drift are outstanding. Combined with the low level of vibrations and high temperature stability of the cryostat, this allows precise measurements over periods of many days.A series of group grown nanodiamond samples were characterised at close to 10 K with photoluminescence measurements. Inhomogeneous broadening is characterised and thermal broadening of the spectra as temperature varies allowed for estimates of the nanodiamond temperatures. Active investigation of more home grown samples is ongoing, including looking for germanium vacancy sites, examining variation between nanodiamond samples and looking at the cooling performance of nanodiamond samples in this cryostat.Mode of access: World wide web1 online resource (xiv, 64 pages) diagrams, graph
Investigating amyotrophic lateral sclerosis candidate genes
No full textEmpirical thesis.Bibliography: pages 121-134.1. Introduction -- 2. Subjects and materials -- 3. Methods -- 4. Reults -- 5. Discussion -- Appendices -- Abbreviations -- References.Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease of motor neurons.Approximately 10% of cases are classiffed as familial and 90% sporadic, although a proportion of sporadic cases have an unrecognised family history. This thesis aimed o identify novel ALS genes from familial and sporadic cases. Whole-exome and whole genome sequencing analysis identified 22 candidate gene variants in an ALS family,which were investigated using an in silico analysis pipeline including protein predictions,conservation, genic tolerance, tissue expression and gene function. Three top priority variants, in CYB5R3, DCAF7, and SAV1, were assessed using in vitro toxicity and localisation assays, with one candidate, CYB5R3, displaying a differential cytoplasmic protein expression pattern. Whole-genome data from 635 sporadic ALS patients were interrogated for ALS candidate genes (n = 21), and ten novel variants were identified, including a novel CYB5R3 variant, which rated strongly from in silico tools. Lastly, this project sought to develop a CRISPR-Cas9 protocol to generate ALS cell models for future use in the in vitro analysis pipeline. Overall, a combination of genetic, in silico and in vitro pipelines were used to identify candidate genes in an ALS kindred and apparently sporadic patients. Novel ALS genes will further our understanding of disease biology and contribute to development of diagnostics and treatments.Mode of access: World wide web1 online resource (xviii, 134 pages) colour illustration
Intelligence-led policing: interpretation, implementation and impact
No full textTheoretical thesis.Bibliography: pages 242-262.Chapter 1. Introduction -- Chapter 2. Review of literature -- Chapter 3. Methodology -- Chapter 4. Policing context in the Maldives -- Chapter 5. Perception of intelligence-led policing -- Chapter 6. Similarities in practice and rhetoric -- Chapter 7. Factors influencing intelligence-led policing -- Chapter 8. Conclusion and recommendations -- References -- Appendices.The resurgence of intelligence-led policing (ILP) in policing policies and scholarship after the 9/11 attacks saw the introduction of new models and interpretations of this philosophy, which added more ambiguity to an already nebulous construct. This research is motivated by this conceptual ambiguity. The objective of this thesis is to analyse ILP through studying its interpretations, implementation, and the perceived impact on policing. The research used the Maldives Police Service (MPS) as a case study.The analysis revealed that the MPS misinterpreted ILP as operating the agency by its intelligence function (sector) by collecting information covertly and disseminating those for tactical operations or supporting investigations (evidence gathering). The MPS held such a narrow definition of ILP due to the lack of adequate effort to educate its staff on ILP, and a failure to introduce the necessary changes to its training materials. The impact of ILP in the MPS was found to be minimal from the perspective of practitioners as: (1) a majority believed that ILP was mainly being used as a tool against political opponents rather than in actual policing, and (2) intelligence generated was largely unactionable. However, in a real sense, the drive to ILP brought a huge change in the MPS as the generation of intelligence products became a regular practice only after the move towards ILP.The ILP literature examined for this thesis pointed to three broad problems. First, the definitional, conceptual and theoretical ambiguity of ILP has made the ILP philosophy not only nebulous for practitioners but also academics and policymakers. Due to this, many agencies adopted the ILP moniker but not the philosophy. Second, National Intelligence Model, the most embraced version of ILP, was too complex for many agencies, including the MPS, encumbering implementation and practice. Third, excessive emphasis on technology in ILP resulted in many agencies becoming fixated with acquiring technical gadgetry, ignoring other important aspects. Consequently, ILP had little positive impact on operational policing in agencies that adopted it.In order to alleviate these problems, this thesis suggests co-practising other policing philosophies with ILP to maximise efficiency and promote democratic values. The American Model of ILP prescribes this principle, but does not delineate it in a way it can be operationalised. This thesis proposes a new ILP model which consolidates the American Model by creating a framework that can be operationalised by any agency based on its context and need.Mode of access: World wide web1 online resource (xi, 284 pages) diagrams, table