Filters
Availability
Publication Year
Resource Type
Faculty
School/Institute
Publication Search Results
Sort
Results per page
1 - 10 of 576
-
(2012)Journal ArticleThis essay is based on the seventeenth annual D. F. McKenzie lecture given at the University of Oxford on 2 March 2011. McKenzie is best known for his tolerant unrolling of the bibliographical into wider cultural and social realms, together with his implication that the material book can be treated as an index of them. Eggert’s essay argues for a timely counterbalance: the need for a regathering of the concept of the work, which largely fell from view during the period of high literary theory. A new model for interpretation is proposed by building reading, and thus the aesthetic, into the definition of the work, while retaining the evidentiary witness of the material book. The model offers, in other words, a justification for folding book history and bibliography into the literary study of all works that have enjoyed an extended history of printing and reprinting.
-
-
(2014)ThesisThe concept of model consistency is espoused as a guiding principle to balance the science and art of economic modelling. Striking a delicate balance between the theory and art of modelling is examined as a way to promote economics as a social science. This is accomplished by looking at more traditional approaches to economics, especially those encapsulated in consumer theory and the theory of the firm, and asking the question: â How can research based on these types of theory on the one hand be â scientificâ and on the other hand be responsive to â socialâ orientations such as the welfare of consumers and the profitability of firms?â A review of a diverse strand of literature shows that the significance of the methodology of model consistency has already been recognised in the macroeconomic context. However, application to the microeconomic context is rather limited. The approach is analogous to â measurement with theoryâ and may be seen as a response to the challenge that â measurement without theoryâ is as useful as â theory without measurementâ . It may also be seen as a way to address the Lucas Critique; a way to bring together theory and application, and as a good complement to experimental economics. The approach is quite suitable for consumer theory and producer theory alike. This is particularly so in a static decision-making context, where producer theory may be considered as a mirror image of consumer theory and the â atemporalâ economic problem of the firm can be analysed using similar techniques as those employed in consumer theory, especially in the single output case. Building upon the logical connection between the utility function and the single output production function, the ideas of consumer theory are carried over to the producer case through an appropriate use of duality theory. This thesis exploits the similarity between the unobservability of utility and the unobservability of â outputâ in a multi-output setting. Duality theory provides the desired rigour of economic science. The quality of its operationalisation is dependent upon the art of modelling. An appropriate mix of price index specification and statistical techniques is required to achieve the desired objectives.
-
(2014)ThesisHyperspectral data provides rich information and is very useful for a range of applications from ground-cover types identification to target detection. With many benefits they also present some challenges including high storage cost, intensive computational load and difficulties in machine assisted interpretation, namely, in classification. The limited number of training samples may cause a significant loss in classification accuracy. This thesis investigates effective and feasible approaches to reduce the dimensionality of the hyperspectral images while keeping the intrinsic structure of the input data intact. The first study is concerned with finding a subspace which consists of the most informative features for reliable hyperspectral image classification. In this study, a hybrid approach which combines both feature extraction and feature selection is proposed. Principal Component Analysis (PCA) is applied first to generate new features from the complete set of the original spectral bands. Feature selection is then performed effectively using a normalized mutual information measure with two constraints to maximize the general relevance and minimize redundancy to the target class identification in the selected subspace. Improvement of the existing nonlinear feature extraction method is undertaken in the second study. In this study, the input features are decorrelated at the first step by applying nonlinear kernel principal component analysis. The spatial properties of the input features are then incorporated to select a subset of features which better reveal object structures and provide good separation among the classes of interest. The third contribution of this study is the evaluation of a number of recent approaches for kernel selection and an improved and computationally efficient approach is proposed. The alignment between the target kernel matrix and input kernel matrix is used to select the kernel parameter(s) for each candidate kernel function. Cross-validation is used at the final stage to search for the best kernel function using the selected kernel parameter(s) for each function. Experiments were carried out on both real and synthetic data. The results show that the proposed approaches provide an improved classification performance.
-
(2018)ThesisThis thesis traces the origins and evolution of the ideas and concepts associated with the revolution in military affairs (RMA) with Chinese characteristics. More specifically, it identifies the group of RMA enthusiasts, tracing their patterns of activities, identifying their communication platforms and channels of influence, and examining their long-term impact on the RMA and ideas associated with it. It argues that from 1980 to 2002, the policy ideas created and advanced by this RMA epistemic community (RMA EC) were crucial in defining PLA conceptions of the RMA. In the early 1980s, they contributed to the reassessing of the international security environment and shaped the Chinese leadership’s threat perception which eventually led to the shift of PLA strategic thought from preparing for imminent all-out war to peacetime army building. They also advocated a holistic, forward-looking approach to defence studies. In the mid-1980s, they proposed major PLA-wide future war studies initiatives, which resulted in introducing the concepts of local war and high-tech wars into the PLA. This eventually led to the strategy of ‘local war under high-tech conditions’, announced in 1993. In the 1990s, they kept expanding the RMA EC and engaged with military regions and group armies, disseminated their future high-tech war ideas to combat units and helping them create operational concepts. In the late 1990s and early 2000s, they played a leading role in identifying the information aspect of warfare as the key of future high-tech wars. This made a significant contribution to updating the PLA’s strategic outlook from ‘local war under high-tech conditions’ to ‘local war under conditions of informationisation’. Moreover, they were among the first to introduce foreign advanced training methods such as computer simulation, realistic combat training and base-ised combat training. In sum, the RMA EC played a leading role in introducing foreign military ideas and adapting them to Chinese contexts, which eventually defined the RMA with Chinese characteristics.
-
(2018)ThesisAdaptive optics (AO) has been proved to be an invaluable technique to improve the imaging through turbulence for ground-based telescopes. By measuring the wavefront distortion in the input beam and consequently correcting in real time, modern large telescopes are able to approach their diffraction limits. One of the difficulties limiting the common adoption of AO for small size telescopes arises from the requirement of wavefront sensing at a high speed to match the temporal bandwidth of the turbulence and its associated complexity in computation and cost. The holographic modal wavefront sensor (HMWFS) promises a simple, high speed and low cost wavefront sensing scheme, where the modal contents of the input aberrations can be measured directly by calculation of intensity ratios between the pairs of spots reconstructed from the hologram. It has attracted much attention, due to these benefits, from the research communities in the past 10 to 15 years, but mainly discussed under numerical simulation and limited laboratory test under coherent laser sources. To explore its usability in AO systems for astronomy, we demonstrated the necessary improvements from an engineering perspective. An improved detection method was proposed to improve the response from the HMWFS, with dedicated hardware implementation to guarantee the detection speed. Common holography techniques are applied to the construction of HMWFS to better meet the foreseen challenges of hologram reconstruction under wide spectrum and low photon star light. Field-programmable gate array (FPGA) devices have been regarded as an ideal platform to implement the heavy computation tasks in the wavefront sensing and reconstruction processed due to its powerful parallel processing capabilities. In this thesis, we developed a stream-based centre of gravity (SCoG) algorithm to estimate the centroid of spots and track their movements automatically. The algorithm was mainly evaluated in a Shack-Hartmann wavefront sensor due to its superiority in terms of reducing the noise errors and expanding linearity and dynamic range. However, it is general enough to be extended for other applications such as supporting the HMWFS or measuring the tip/tilt by tracking the focal spot. An FPGA implementation of the SCoG algorithm has been tested both in laboratory experiments and on-sky observation to validate its usability.
-
(2018)ThesisThe period of democratic transition in Myanmar, beginning in 2010, has seen the emergence of devastating communal violence, the vast majority of which has been directed towards Muslim communities. This thesis considers the phenomenon of communal violence during democratic transition in Myanmar through a contentious politics framework. The historical institutionalist argument made in this thesis suggests three factors necessary for communal violence in this case; an exclusive definition of the political community at the time of democratic transition, elites’ promotion of exclusionary forms of nationalism during democratic transition, and the availability of non-elites willing to perpetrate violence. This thesis documents a mechanism between democratic transition and exclusionary forms of nationalism in Myanmar. The practice of democracy requires the definition of a nation’s boundaries – of who is a member of the political community and of who is not. Democracy, however, cannot determine where these boundaries should lie. It is instead nationalism which will define a new conception of the political community. This thesis finds that the exclusionary forms of nationalism which arose during democratic transition in Myanmar reflected the historically exclusive definition of the nation and the construction of particular communal groups, foremost the Rohingya, as a threat to it. Such conceptions were promoted by political, religious and other elites during transition. Using a dynamic contentious politics approach to study the interactions of actors, opportunities and mechanisms, this thesis stresses the ways in which communal violence interacted with democratic transition. This is considered through an analysis of the processes of; the historical construction of national and communal identities, the activation of the boundaries of these identities at democratic transition, the constitution of actors (including political parties and nationalist organisations), brokerage between these actors, and the interpretation of violence by the state and other actors as democratic transition continued. Insecurity is considered as an environmental mechanism which influences the attribution of threat by actors. Mobilisation for violence can be found throughout these processes and mechanisms. This thesis draws upon in-depth interviews with political, religious and community leaders and other community members, predominantly taken in five different case-study locations across Myanmar.
-
-
(2017)ThesisCurrent design standards do not appropriately recognise the characteristic nonlinear stress-strain behaviour with significant strain hardening offered by stainless steel. Loss of effectiveness due to local buckling is dealt with effective width method, which is unjustified for stainless steel as there is no obvious yield plateau. With recent developments, the continuous strength method (CSM) was shown to accurately predict cross-section resistances of stocky sections exploiting the benefit of strain hardening. The current research extended the scope of CSM for slender sections retaining the same base curve, and also develop CSM based design guidelines for predicting the buckling resistance of stainless steel columns. A new concept of using the equivalent elastic deformation capacity, defined as the elastic strain at ultimate load, was proposed for slender sections. A comprehensive FE study was carried out to establish relationships between the equivalent elastic deformation capacity and buckling strain of CSM. Proposed relationships allowed using the same CSM base curve for both stocky and slender sections, and hence, cross-section resistances could be directly determined using CSM buckling stress and gross cross-sectional properties. Once CSM was successfully implemented for all cross-section types, the current study extended its scope to the member level. A series of new buckling curves was proposed for cold-formed RHS and SHS columns using available test results and generated FE results. The proposed technique combined CSM buckling stress with Perry type buckling curves. Once the proposed buckling formulas performed well for cold-formed hollow sections, the flexural buckling of welded I-columns were thoroughly investigated through experimental and numerical methods. Valuable experimental evidences on column buckling of welded sections were added to currently available limited data. An extensive FE analysis supplemented the test results, and all available results were used to develop CSM based buckling curves for welded I-sections considering the effects of residual stresses and other important parameters. All developed formulas were shown to produce accurate, consistent and reliable predictions.
-
(2017)ThesisSpares determination is the process of determining what spare parts, in the right quantity and when they are required, to support the maintenance of a physical asset such as a ship. With an understanding of the configuration and mean time between failures of an asset’s components, organisations are able to determine the maintenance and parts required to keep it at the required level of availability. Feedback through use should further enhance asset knowledge leading to more efficient and effective spares determination. The thesis presents evidence that organisations are ignoring theory in acquisition and feedback loops in asset operation, rendering their spares determination inefficient and ineffective. A significant gap in spares determination knowledge was found that indicated that factors such as culture, leadership and demographics (human factors) may explain why the theory was not being applied. The project used a mixed methods approach. The quantitative phase consisted of a workforce survey conducted across 14 different business units (n=219) in the Australian Department of Defence. Responses were analysed through multiple regression with spares determination performance statistics. Statistical models supported the proposition, with approximately 50% of the sample’s variation explained by human factors. Surprisingly, the models indicated that theory was being applied to drive effectiveness not efficiency as it was designed. These unexpected results were explained by the qualitative phase of the project, which included a total of 25 semi-structured interviews of personnel from four business units. These found dominant military ‘can-do’ culture which emphasised effectiveness to drive asset availability at the expense of efficiency. Where transformational leadership existed, entrepreneurial behaviours where modelling, experimentation and planning provided more effective spares determination. These business units also overcame a series of organisational issues preventing better performance. The thesis quantified the large impact of human factors on spares determination performance. It found that organisational culture and transformational leadership were the factors that had the greatest influence on spares determination efficiency and effectiveness. These findings provide an opportunity for further theoretical and practical development of asset management and integrated logistics support that should yield a substantial increase in spares determination performance.