Sydney eScholarship
Not a member yet
31878 research outputs found
Sort by
The locality of quantum codes
No full textThis thesis concerns itself with the role locality plays in the performance of quantum codes.
We first address the difficulty of producing codestates from relatively good codes, specifically we
investigate the depth of geometrically local encoding circuits. Previous results struggled to produce
bounds allowing for general, non-local classical communications in the circuit. We show that, owed to
the entanglement hidden in codestates, the depth of encoding circuits is lower bounded, even when
allowing for boundless classical computation.
We then move on to address a converse question: given a particular code, what can we say about its
geometrical structure? We show that implementing stabiliser codes requires satisfying some
connectivity requirements that depend on k and d. We quantify these requirements through the
notion of graph expansion.
Finally, in the last chapter, we ask the question of the achievability of these requirements. The
celebrated BPT bound states upper bound on the parameters that can be achieved for codes local in
three dimensions. We provide the first construction that matches this bound
Women in profit-to-member superannuation funds
No full textA research report examining the representation of women in leadership and management positions in 59 Australian profit-to-member superannuation fund
Geometric extensions and the six functor formalism.
No full textThe results in this thesis are linked by their use of the six functor formalism. In the first chapter, we
introduce geometric extensions, canonical sheaves on singular varieties characterised by their
occurrence as a summand of the cohomology of any resolution of singularities. These objects
generalise intersection cohomology, parity sheaves, and provide a definition of intersection K-theory.
In the second chapter, we interpret this construction in the context of real algebraic varieties. This
leads to a real interpretation of the mod two Hecke category, and supplies a definition of mod two
intersection homology groups on real algebraic varieties, answering an old question of Goresky-
MacPherson. In our third chapter, we give a string diagrammatic interpretation of various maps in the
six functor formalism. This graphical calculus leads to the proof of a general coherence theorem.
While this theorem does not incorporate the monoidal aspects of the theory, it gives the first
coherence result in a six functorial context that treats all four induced functors, along with the natural
transformations between them
Toward A Generalised Signal Processing Framework for Inter-Subject Associative BCI
No full textAn inter-subject associative BCI refers to a fully zero-training neural decoding algorithm when a group of users shares similar EEG features so that the algorithm can perform well for a user when trained on other users' data from the same group. However, only a few studies attempted to quantify inter-subject associativity, i.e., identifying task-related EEG features that exhibit unchanged feature domains across subjects, minimizing the chance of covariate shift occurrence. Quantifying inter-subject associativity complements data-driven transfer learning and delivers good BCI performance without any training data. Data-driven conventional techniques, including common spatial pattern (CSP), are prone to overfitting and often demonstrate inconsistent classification accuracies. However, recently introduced convolutional neural network (CNN)-based architectures are deep learning techniques that inherently learn features from data with nonlinear activation functions while optimizing the models' predictive capabilities. A novel Bhattacharya distance-based predictor was developed for a CSP-based BCI classification framework. The CSP-based methods were compared with novel BCI classification pipelines utilizing a 1-dimensional convolutional neural network (1D-CNN) architecture. Various feature representation techniques, such as bandpass-filtered EEG signals, power spectral density (PSD) sequences, and bi-channel cross-power spectral density (CPSD) sequences, were used to train the proposed 1D-CNN architectures. The proposed methods were tested on motor imagery (MI) and speech classification tasks from EEG signals. Results implicated that 1D-CNN, utilizing time-domain EEG signals, produced better classification accuracies than frequency-embedded PSD or CPSD sequences and CSP-based methods for intra- and inter-subject MI classification. However, the proposed 1D-CNN with PSD sequences outperformed the results of time-domain EEG signals for intra-subject speech BCI
Machine learning-assisted lipidomics informs personalised prediabetes outcomes following lifestyle interventions
No full textPrediabetes is among the most rapidly growing global health challenges, affecting over 720 million individuals worldwide. Lifestyle modifications aimed at weight loss are the primary strategy for prediabetes management, offering glycaemic benefits, particularly in those with excess body weight. However, significant inter-individual variability in glycaemic responses and the risk of weight regain during long-term lifestyle interventions present major challenges, compromising effective prediabetes management. Currently, there is no established biomarkers to evaluate effectiveness of lifestyle interventions or to predict who are less likely to achieve glycaemic improvements. This critical gap restricts timely and personalised therapeutics, increasing the risk of type 2 diabetes and its complications.
Emerging lipidomics analyses revealed that shifts in circulating lipid profiles contribute to insulin resistance, β-cell dysfunction, and microbiome imbalances, all of which modulate prediabetes progression or reversion. Key lipid species, including signalling sphingolipids, pro-inflammatory lysophospholipids, nutrient-related glycerolipids, obesity-related glycerolipids, microbiome-derived short chain fatty acids, and metabolically active free fatty acids, have been widely implicated in prediabetes regulation. However, their potential as evaluative or predictive biomarkers for managing overweight/obesity-associated prediabetes through lifestyle interventions remains largely unexplored.
This thesis leveraged high coverage lipidomics to analyse serum samples from the Australian sub-cohort of the PREVention of diabetes through lifestyle Intervention and population studies in Europe and around the World study. The aim was to identify lipid biomarkers capable of evaluating or predicting glycaemic responses to an 8-week low-energy diet-induced acute weight loss (Project 1) and subsequent 3-year weight-maintenance interventions combining diet and physical activity (Project 2)
Organic Synthesis on Strigolactones Chemistry
No full textStrigolactones (SLs) are carotene-derived natural products that play important roles as phytohormones (within plants), as seed germinators of parasitic weeds (between plants), and as chemical communicators between plants and fungi (between kingdoms). A subset of these molecules, called non-canonical SLs, exhibit highly diverse structures that present significant challenges to the synthetic chemist.
The research topic of this thesis is focusing on developing novel synthesis routes from commercially available ionones to two strigolactone molecules: carlactone (CL) and avenaol. Based on the knowledge gained en route to CL, further research was conducted towards CL derivatives which shared the same carbon skeleton as CL.
The choice of which non-canonical SLs to target for synthesis was not random. As the biosynthetic precursor of all SLs, and a natural product in its own right, the importance of CL is self-evident. However, total synthesis routes to this molecule remains relatively under-explored. There is still no large-scale preparation that can produce enough CL to enable further biological research. Therefore, inventing a cost-effective and chemoselective synthetic route is necessary.
Avenaol is perhaps the most synthetically challenging non-canonical SL, possessing a unique fused cyclopropane unit that is incredibly sterically congested. We have designed a novel synthetic strategy that intercepts an intermediate in the only reported synthesis of avenaol. We aimed at significantly simplifying the total synthesis of avenaol, thereby facilitating material access for further studies into its potential biological or therapeutic properties
Urban Displacement in New York City and Gentrification as City-wide Strategy, 2019-2024
No full textThis thesis examines the causes and catalysts of urban displacement in contemporary New York City. More than a physical relocation urban displacement is understood as a spatial and ontological effect of intertwining socio-spatial forces driven by neoliberal economic practices. As the commodification and financialisation of housing and urban space intensifies gentrification becomes a city-wide strategy rather than a neighbourhood-bounded phenomenon. This research highlights the complexities of the reproductions of social spaces that result in urban displacement drawing on Lefebvre’s concepts of the spatial triad and abstract space to demonstrate how gentrifying urban space is produced and maintained. By analysing changes to city and state policies aimed at addressing the city’s housing crisis – before during and in the shadow of the COVID-19 pandemic – this research shows how the political and legal strategies its residents rely have shifted. The research is primarily concerned with renters who given the history of uneven development in the city are more vulnerable to displacement and the structural forces behind gentrification than homeowners. As displacement extends beyond the home public space is also at stake. Differing political conceptions of how to maximise the use value of urban spaces pit competing groups of local actors against each other. This research shows how the local is heavily dependent on the political context within a social space. Resistance to urban displacement becomes essential in city-wide movements for spatial justice and this research shows how resistance is always also politically and spatially contextual and justice itself is an ongoing process never completely realised. Through fieldwork and an in-depth study of policy law and media this research highlights how political and legislative wins secured by housing justice actors contain imperfections and openings through which institutional actors can perpetuate displacement
Animal Communication in Hyper-Contemporary Australian Literature: Genre, Ethics, and Climate Change
No full textThis project investigates how genre conditions the representation of animals in hyper-contemporary Australian fiction engaged with the environmental devastation of climate change. Analysing five novels by settler authors - Robbie Arnott’s 'Limberlost' (2022), Erin Hortle’s 'The Octopus and I' (2020), Charlotte McConaghy’s 'Once There Were Wolves' (2021), Laura Jean McKay’s 'The Animals in That Country' (2020), and Chris Flynn’s 'Mammoth' (2020) - I examine how these works navigate the representational and ethical complexities of the Anthropocene by centring animal presence and perspectives. This thesis traces how writers employ a spectrum of generic modes, from realism to speculative fiction, to depict animals and their interactions with humans, demonstrating how diverse genres convey animal presence and agency. Crucial to this research is my theorisation of ‘animal communication,’ a concept that transcends anthropomorphism by exploring how writers’ representations of gaze, behaviour, voice, and human-animal interactions depict animal presence and subjectivity without relying solely on animal interiority. This thesis challenges Amitav Ghosh’s critique of realism, demonstrating how the realist mode, alongside speculative and experimental approaches, addresses the vast, nonhuman dimensions of climate change effectively. Furthermore, it examines critically settler-colonial legacies embedded in these narratives, focusing on their portrayal of violence, care, and ethics in human-animal relationships on colonised land. By revealing how genre shapes the representation of animal subjectivity and environmental ethics through animal communication, this thesis offers a nuanced perspective on contemporary Australian literature at the intersection of environmental ethics, animal studies, and literary form. Its findings contribute to broader discussions of the power of storytelling to reimagine human-animal relationships and inspire ethical engagement in the Anthropocene
Constructing City Images in Short Videos: A Comparative Study of Beijing, Xi'an, and Shanghai
No full textThis study examines how short videos facilitate the construction of city images through case studies of Beijing, Xi’an, and Shanghai, and aims to reveal both shared patterns and distinctive strategies in the three cities’ image construction. Anchored in the city image communication theory and film aesthetics, it adopts the qualitative content analysis and comparative research, delving into the five representational techniques: color, light and shadow, sound, characters, and post-production.
The research findings show that creators’ choices are shaped by a city’s resources, development stage, and communication positioning, resulting in varied strategic emphases in each city’s video cases. Beijing’s short videos highlight solemnity and historical depth. Xi’an’s short videos underscore traditional culture and everyday vitality. Shanghai’s short videos blend modern dynamism with individual emotions. Meanwhile, as influenced by platform algorithms and the operational rationale of short videos, the five representational techniques operate in tandem. Color, light, and shadow jointly set the visual tonality. The sound system guides the delivery of emotions and rhythms. Characters serve as narrative anchors and experiential vectors. Post-production techniques organize and amplify the information expression, enhancing immersion and memorability of the viewing experience.
This study delivers academic contributions in three dimensions. Theoretically, it incorporates micro-level representational techniques into the digital communication framework, enriching the perspectives of city image research. Methodologically, it proposes a multi-dimensional analytical model tailored to the short video context. Practically, it offers actionable insights into future city branding and content creation. Overall, the study fully reveals how cities are constructed as both symbolic and experiential images in the contemporary Chinese media environment
Deep Learning-based Tumour Heterogeneity Analysis with Multiparametric Magnetic Resonance Imaging
No full textTumour heterogeneity, characterised by spatial and temporal variations in cellular morphology and molecular profiles within the tumour tissues, impacts oncology diagnosis, prognosis, and treatment outcomes. Medical image analysis, particularly through multiparametric magnetic resonance imaging (mpMRI), provides a non-invasive approach to capture these complex heterogeneity patterns, enabling a more objective and comprehensive assessment of tumour biology. However, the intrinsic variability in heterogenous tumours poses challenges in accurately delineating tumour regions and predicting therapeutic responses, often leading to inconsistent clinical interpretations. Because of this, labelling and analysing heterogeneity sub-regions in mpMRI are time-consuming tasks and requires experienced expertise.
This thesis introduces a novel deep-learning framework that addresses the challenges of tumour heterogeneity in mpMRI modalities to enhance tumour heterogeneity analysis in medical image and its use in downstream tasks including image segmentation and classification. The framework comprises of two main components. First component is an unsupervised semantic segmentation method developed to delineate tumour sub-regions automatically. This method effectively captures the intrinsic structure of heterogeneous tumours by leveraging a multi-phase training strategy that combines coarse segmentation with refined, self-supervised learning enhanced by sparse spatial continuity and context-based hierarchical loss functions. Second, we propose a heterogeneity-aware deep learning method for tumour classification that integrates machine-generated sub-region labels with dual-stream feature extraction for both local heterogeneity and global image information. A learnable alignment module is employed to standardise sub-region labels across different imaging modalities, enabling the extraction of both local heterogeneity features and global contextual information