Theses

Permanent URI for this collectionhttps://uwspace.uwaterloo.ca/handle/10012/6

The theses in UWSpace are publicly accessible unless restricted due to publication or patent pending.

This collection includes a subset of theses submitted by graduates of the University of Waterloo as a partial requirement of a degree program at the Master's or PhD level. It includes all electronically submitted theses. (Electronic submission was optional from 1996 through 2006. Electronic submission became the default submission format in October 2006.)

This collection also includes a subset of UW theses that were scanned through the Theses Canada program. (The subset includes UW PhD theses from 1998 - 2002.)

Browse

Now showing 1 - 20 of 17703

Engineering Development and Signal Processing Advancements in OCT Angiography: From Custom System Integration to Temporal Domain Denoising
(University of Waterloo, 2026-06-24) Perez Paredes, Andrei Felipe
Optical Coherence Tomography Angiography (OCTA) positions itself as a highly effective, non-invasive technique that provides depth-resolved visualization of vascular structure and function. With a continuously emerging need to transition from static angiography to functional, time-resolved imaging, researchers have identified interconnected challenges. This thesis fundamentally explores two of these challenges: speckle noise and processing latency. Typically, spatial filters used to suppress speckle and denoise images are computationally expensive and act as temporal low-pass filters, destroying the dynamic physiological signals they intend to isolate. This thesis presents the design, implementation, and in vivo validation of a streaming-compatible swept-source OCTA (SS-OCTA) architecture relying on a hardware/software co-design to overcome these limitations. Rather than relying on isolated downstream algorithms, the system described in this research establishes a validated quality baseline starting at the hardware level. The custom 1060 nm MEMS-VCSEL SS-OCT platform developed in this thesis, leverages an adaptive software flyback filter to assess fast-axis position derivatives, actively isolating and discarding corrupted scans prior to contrast processing. Building upon this stationary signal foundation, the thesis introduces Temporal Subband Decomposition and Amplification (TSDA). TSDA operates as a dual-rate infinite impulse response (IIR) filter along the per-pixel temporal axis, decomposing the signal into structural, flow, and high frequency speckle bands. This continuous formulation reduces computational complexity to O(1), bypassing the buffering requirements of discrete Fourier methods and aiming to isolate physiologically driven flow from coherent noise. The integrated hardware/software stack was validated against a microfluidic phantom and an in vivo 14-day-old chorioallantoic membrane (CAM) preparation. An ablation study reported here confirms the TSDA architecture achieves a processing latency within the 10 ms budget. Furthermore, the complete pipeline delivered a Peak Signal-to-Noise Ratio (PSNR) of 27.8 dB against a multi-frame average reference, while yielding statistically significant improvements in Vessel Contrast-to-Noise Ratio (VCNR). By replacing spatial averaging with targeted temporal band isolation, the integrated platform extracts OCTA contrast while preserving the temporal flow signal within the filter passband.
Huge Operators in Holography: BPS Sectors, Matrix Models, and Black Holes
(University of Waterloo, 2026-06-24) Murali, Harish
This thesis explores quantum gravity by studying large-N gauge theories and matrix models. In particular, it focuses on operators whose charges scale as N^2, which we dub huge operators, so that they are heavy enough to backreact on the dual bulk geometry. In the first part, we study protected sectors of N = 4 super Yang-Mills theory, where supersymmetry gives enough control to ask finite-N questions beyond the planar limit. We analyze huge 1/2-BPS operators and show that their exact combinatorics reorganizes, at large N, into matrix models and integrable HCIZ fluid flows. We also study the 1/16-BPS sector relevant for supersymmetric black holes, emphasizing the role of finite-N trace relations and analytic continuation in the number of colors. In the second part, we turn to simpler matrix models as laboratories for holographic ideas such as universality, and commutativity. We show that huge deformations can produce universal eigenvalue densities in strong-coupling regimes, and we clarify the role of fermions in ensuring commutativity at strong coupling. Together, these results give concrete boundary descriptions of backreacted geometries, finite-N effects, and strong coupling dynamics.
Digital divide and financialization in Canada’s rental housing sector: An epistemological critique of technology in urban space
(University of Waterloo, 2026-06-24) St-Hilaire, Cloé
Housing is one of the most physical, tangible components of everyday life, and yet as rentiers occupy more and more space in the economy and society, housing is transformed into abstract financial products and digitally rendered into platforms, apps, and data. The increasing presence of platforms in urban life, primarily under corporate ownership, is reshaping how we view, research, understand, and experience housing. For tenants, it has translated into the digitization of the rental housing experience, from apartment search, tenant screening, to monthly payments. For landlords, it has meant increasing means for extracting value from tenants, derived from insights produced by data. For researchers and activists, it manifests as opaque information landscapes, leaving key questions unanswered and hindering housing justice efforts. For policymakers, it remains a display of fragmented data infrastructures. As housing continues to embody its contradictory nature between home (use value) and profit (exchange value), the digitization of rental housing warrants further scrutiny into how it contributes to the speculative conditions of housing under rentier capitalism. This thesis offers an epistemological investigation of the rise of data and digital technologies in Canada’s rental housing sector. It argues that the deployment of technology in rental housing, and the data that is produced as a result, (re)produces uneven epistemic outcomes that benefit capital and hampers social justice. This digital turn in housing has been led by rentiers who use platforms, apps, algorithms, and data for the production of housing information. By controlling the data pipelines, rentiers are able to dictate what gets measured (and what does not), frame housing digitization under deterministic discourses of progress and efficiency, and limit other’s capacities to know via the corporate gatekeeping of information. This leads to epistemic injustices against those who become targets, objects, and test subjects of housing datafication, and who are at the same time prevented from meaningfully understanding how this datafication affects them. From an urban governance perspective, the city under digitization remains governed through veils of opacity marked by inadequate data infrastructure, also creating epistemic injustices. This analysis combines a qualitative document and media overview of proptech and finance in Canada, a spatial analysis of proptech adoption in the build-to-rent sector of four cities, select Canadian case studies, key informant interviews, and a large-scale analysis of housing data infrastructures. The findings are separated into four empirical chapters pertaining to proptech and/or ownership data. The first article critically examines the socio-technical imaginaries of proptech–efficiency, lifestyle, sustainability, and democratization–as carried through by the industry and the media, and how these imaginaries are examples of technological determinism. The second chapter analyzes the adoption of proptech in Canada’s build-to-rent housing submarkets in Vancouver, Calgary, Toronto, and Montreal, the major adopters of rental proptech, and the characteristics of buildings with high proptech adoption. The third chapter presents the theoretical concept of epistemic engulfment to help make sense of the implications of the rise of proptech propelled by finance on the epistemic regimes of housing, and its implications for housing and urban justice. The fourth chapter analyzes the ownership data infrastructures of 31 cities across North America and Europe to determine why ownership data remains opaque despite increasing digitizing efforts from states and cities, and how ownership data opacity prevents the answering of key urban questions. In its entirety, this thesis offers an epistemological critique of the rise of digital technologies in rental housing under financialization through an analysis of its discourses and data infrastructures. It illustrates how the production of housing data under the control of private actors contributes to the already uneven power relationship between landlords and tenants through injustices that are epistemic in nature. It shows how urban governance contributes to the making of the conditions that allow us to know about urban issues, or remain in the dark. This thesis inserts itself in larger discussions about viewing housing data as a political subject and urges planning scholars to endeavour in critical reflections about urban information.
Groundwater and vegetation influences on alpine wetland evapotranspiration
(University of Waterloo, 2026-06-24) Murray, Eric
Wetlands are increasingly recognized for their ecological significance and hydrological function, particularly in snowmelt-dominated mountain regions experiencing climate change. This thesis investigates evapotranspiration (ET) and groundwater-surface water interactions within the Burstall Wetland, a mineral wetland located on the eastern slopes of the Canadian Rockies. The study aims to (1) examine seasonal wetland-scale ET fluxes and the relative contribution of snowmelt versus growing-season processes, and (2) identify the sub-surface and vegetation controls on spatial ET variability during the snow-free period. Data collection was conducted during the 2022 growing season using eddy covariance (EC) to measure wetland-scale energy and carbon fluxes and a closed dynamic chamber system to capture microsite ET across dominant vegetation communities (sedge, willow, moss, and litter). Groundwater levels were monitored through a network of groundwater wells, and volumetric water content (VWC), soil temperature, and meteorological variables were recorded to support ET estimation and spatial analysis. By integrating site-scale flux observations with chamber-based measurements, this study characterizes the spatial heterogeneity of ET and evaluates the contribution of groundwater to seasonal loss. The findings provide insight into the ecohydrological processes governing alpine wetland function and offer a baseline for assessing wetland sensitivity to future climatic and hydrological shifts in mountain environments.
Towards Trustworthy Federated Learning: Security, Privacy, and Verifiability
(University of Waterloo, 2026-06-24) Deressa, Biniyam
Federated learning enables collaborative model training across institutions that cannot share raw data, but practical deployments rely on trust assumptions that do not hold in adversarial environments. Malicious clients may omit or falsify computation, inject poisoned updates, or free-ride on collective training with negligible detection risk. Existing defenses address security, privacy, and verifiability in isolation: privacy mechanisms obscure the signals required for robustness, while general-purpose zero-knowledge proof systems incur costs that scale with circuit size and are impractical for neural network workloads. The result is a structural \emph{trust deficit} that no single existing mechanism resolves. This thesis argues that the security--privacy--verifiability tension in federated learning is \emph{architectural rather than fundamental}. By decomposing trust into \emph{four separable research problems}, namely, adversarial client selection, privacy-compatible robust aggregation, cryptographic training verification, and compositional architecture, and by exploiting the algebraic structure of learning workloads, each property can be enforced by a mechanism with explicit assumptions and well-defined interfaces. These mechanisms are independently deployable and compose via defined interfaces without requiring cross-mechanism security re-analysis, yielding a \emph{modular trust architecture} for trustworthy federated learning. \textsc{TrustBandit} addresses the security dimension by formulating client selection as an adversarial multi-armed bandit under partial observability. Importance-weighted reputation estimation with adaptive exploration achieves a provable regret bound $O(\sqrt{T N \ln N})$, where $T$ is the number of training rounds and $N$ is the number of clients, and, in evaluation, identifies trustworthy clients with $94$--$99\%$ success in low-adversary settings (up to $20\%$ adversaries) and maintains $67$--$69\%$ selection success under $50\%$ adversarial participation, while sustaining $70.97\%$ test accuracy at $50\%$ adversarial participation and improving robustness by up to $5.5\times$ over standard selection baselines. \textsc{PROFILE} addresses the privacy--robustness tension through architectural separation rather than algorithmic compromise: anomaly detection is relocated from centralized plaintext inspection to server-side predictive detection over bucket-wise homomorphically encrypted aggregates with semantic client assignment. The framework enforces IND-CPA computational privacy for individual updates under Ring-LWE hardness, with LDP-protected metadata, while preserving Byzantine robustness under poisoning and backdoor attacks; empirically it achieves accuracy within 2--3\,pp of the best plaintext baseline (FLTrust) while operating under full RLWE encryption, with detection rates from $0.87$ to $0.99$ across all datasets and non-adaptive attack types; adaptive adversaries that suppress per-round statistical signals fall outside this bound, as characterised by the leakage--detectability frontier. \textsc{zkMaP} and \textsc{zkExp} address verifiability by specializing to the dominant computational kernels in training. \textsc{zkMaP} gives succinct verification for matrix multiplication via polynomial identities over pairing groups, achieving $O(n^2)$ prover complexity for matrix dimension $n$, constant-size proofs (320 bytes), and constant-time verification (3.68\,ms), yielding up to $19.07\times$ verification speedup over prior specialized matrix multiplication protocols at comparable security. \textsc{zkExp} provides a succinct proof system for exponentiation with constant-time verification and constant-size proofs (160 bytes for single proofs; 256 bytes in batched mode), with low amortized batch overhead (1.35$\times$). \textsc{RIV} composes these primitives into an end-to-end proof-of-training protocol. Training transcripts are committed prior to challenge selection, preventing selective honest computation. Stochastic Interval Commitments certify native IEEE-754 floating-point computation within backward-error-derived bounds while preserving cryptographic binding. The resulting protocol provides parameterized detection guarantees: for an adversary corrupting a $q_{\mathrm{adv}}$-fraction of challenged layers, the per-round acceptance probability is bounded by $(1-q_{\mathrm{adv}})^k + k\varepsilon_{\mathrm{crypto}} + \delta_{\mathrm{fp}}$ (where $\varepsilon_{\mathrm{crypto}} \le 2m/|\mathbb{F}_p| + \mathsf{negl} \lambda)$ per challenged layer), yielding explicit trade-offs between challenge rate, overhead, and adversarial detectability (e.g., $>99.99\%$ cumulative detection at $k=3$ over 50 rounds). Collectively, these results demonstrate that cryptographically grounded trust in federated learning is achievable through specialized, composable mechanisms rather than monolithic designs.
Sustainability Management in Private Capital Markets: Important and Distinct, yet Underexplored-Institutional Pressures, Legitimacy, and ESG Disclosure
(University of Waterloo, 2026-06-24) Mirza, Majid
The rapid expansion of sustainable finance has intensified demands for consistent sustainability disclosure across global capital markets. While public market actors and listed corporations have received significant scholarly attention, private capital markets, particularly private equity, remain comparatively underexamined despite their growing influence in global investment flows. This dissertation investigates how sustainability management and disclosure practices are emerging within private capital markets and how private equity actors respond to evolving institutional pressures shaping environmental, social, and governance (ESG) reporting. Drawing on institutional theory, legitimacy theory, and sustainability management literature, the dissertation explores three interconnected dimensions of sustainability integration in private capital. The first paper presents a systematic literature review of sustainability research within private capital investing, identifying a substantial gap between the rapid growth of ESG practices in industry and the limited academic attention devoted to sustainability within private equity and venture capital research. The review reveals that sustainability-related scholarship constitutes a very small proportion of the broader private capital literature and highlights several emerging thematic areas requiring further investigation. Building on this foundation, the second paper examines ESG reporting practices among leading global private equity firms through a comparative analysis of ESG reports and Sustainable Development Goal (SDG) integration strategies. The findings suggest that while sustainability commitments are increasingly communicated within ESG disclosures, much of the integration appears to function as legitimacy signaling rather than deeply embedded investment decision-making processes. The third paper extends the analysis to the evolving institutional landscape of global sustainability disclosure by examining comment letters submitted by financial institutions and private equity firms in response to the International Sustainability Standards Board’s (ISSB) consultation on agenda priorities. Using a mixed-method approach combining thematic interpretation and structured content analysis, the study identifies patterns of institutional isomorphism and dissonance within the consultation process. While financial institutions and private equity actors demonstrate convergence around biodiversity and climate–nature disclosure priorities, significant divergence emerges regarding the role of integration in sustainability reporting, reflecting distinct institutional logics within the financial sector. Taken together, the findings illustrate how private capital actors both conform to and shape emerging sustainability management frameworks in the form of selective institutionalization. The dissertation contributes to scholarship by expanding understanding of sustainability integration within private capital markets, highlighting the role of institutional pressures in shaping ESG disclosure practices, and introducing private equity as an important and distinct, yet underexplored actor in global sustainability reporting debates.
Quantifying Structural Uncertainty in Hydrologic Models
(University of Waterloo, 2026-06-23) Arabzadeh, Rezgar
Hydrologic models are essential tools for understanding watershed processes and supporting water resource management. However, their predictions are inherently uncertain due to imperfect model structures (structural uncertainty), parameter estimation challenges (parameter uncertainty), and limitations in observational data and model forcings (input uncertainty). Bayesian inference has become a widely used framework for quantifying these uncertainties because it enables probabilistic parameter estimation and prediction while formally incorporating prior information and observational evidence. Despite these advantages, the application of Bayesian methods to complex hydrologic models remains computationally demanding, and the resulting predictive uncertainty often represents a combination of multiple uncertainty sources (including input, parameter, and structural uncertainties) that are difficult to interpret individually. These limitations reduce the effectiveness of Bayesian uncertainty analysis as a diagnostic tool for improving hydrologic models. This thesis develops methodological advances to improve the efficiency and interpretability of Bayesian uncertainty quantification in hydrologic modeling. The research focuses on two challenges: improving the computational feasibility of Bayesian inference for complex models and separating the sources of uncertainty represented within Bayesian predictive distributions. To address these challenges, new methods are developed and evaluated using both regional and continental-scale hydrologic datasets. 1. A machine learning–assisted framework is developed to improve the efficiency of Bayesian joint inference for hydrologic models. The proposed approach integrates machine learning techniques with Bayesian calibration to facilitate exploration of complex posterior parameter distributions and reduce the computational burden associated with traditional sampling methods. The framework is evaluated using twelve watersheds from the MOPEX dataset and demonstrates improved inference performance while maintaining reliable uncertainty quantification. 2. A variance decomposition methodology is introduced to identify and quantify the sources of uncertainty embedded within Bayesian predictions. While Bayesian calibration provides probabilistic estimates of model outputs, it does not directly attribute predictive uncertainty to individual components of the modeling framework. The proposed method decomposes posterior predictive uncertainty into interpretable components, enabling a clearer understanding of how different aspects of the modeling process contribute to overall uncertainty. 3. The proposed uncertainty decomposition framework is applied to a large-scale hydrologic analysis across approximately 3,000 watersheds in North America. This continental-scale application enables the systematic evaluation of spatial patterns in hydrologic model uncertainty and reveals how dominant uncertainty sources vary across hydroclimatic and physiographic regions. Together, the contributions of this thesis improve both the computational efficiency and the interpretability of Bayesian uncertainty estimates in hydrologic modeling. The proposed approaches provide tools for diagnosing uncertainty sources and evaluating model reliability, which can support more transparent hydrological predictions across a range of environmental and water resource applications.
Contributions to the model theory of algebraic differential equations
(University of Waterloo, 2026-06-23) Eagles, Christine
This thesis deals with semiminimal analyses of finite rank types, primarily in the stable theory of differentially closed fields of characteristic zero (DCF0). The two main themes considered in this thesis are determining when a type is minimal or semiminimal, and understanding what invariants of finite rank types are captured by a semiminimal analysis. In DCF0, a central concern of this thesis is determining when a type is almost internal to the field of constants. Partially generalising a result of Rosenlicht, algebraic criteria are provided in two different contexts: rational vector fields on affine n-space, and pullbacks under the logarithmic derivative of certain types which are internal to the constants. The criteria in the former case answers a question posed by Freitag, Jaoui, Marker and Nagloo about when the Poizat equations are internal to the constants. In both cases, the theory of binding groups in stable theories plays a significant role. Results of Duan and Nagloo are improved upon to completely classify when the generic types of Lotka-Volterra systems are minimal. In the minimal case, a characterization of the possible relations that may exist between solutions of distinct Lotka-Volterra systems is given. In the general setting of a totally transcendental theory, it is shown that the multiplicity with which a minimal type arises in a semiminimal analysis of a finite rank type is invariant, i.e., it is independent of the semiminimal analysis. A conjecture is proposed for the possible ways for two semiminimal analyses of the same finite rank type to differ. Along the way, the connection between semiminimal analyses and domination decompositions, is clarified.
Parameter Inference and Model Selection for Differential Equation Models with Applications
(University of Waterloo, 2026-06-23) Zhao, Yuxuan
Dynamic systems are commonly modelled by differential equations (DEs) in epidemiology and biology, among other fields. The parameters in the DEs are often of scientific interest and required for estimation, given a set of noisy observations. The first and oldest general class of methods for the parameter inference problem in DEs is based on numerical solvers. As a preliminary study, we conduct a comparative study of compartmental models for COVID-19 transmission using such numerical solver-based methods. However, this class of methods can be computationally intensive and may only converge to the local optima due to the sensitivity of the numerical solution to the parameters and initial conditions. This thesis begins by presenting this study, which highlights these limitations and motivates the methodological developments that follow. To address these challenges, Gaussian process-based methods serve as an alternative that bypass the need for numerical solvers. In particular, the recent manifold-constrained Gaussian process inference (MAGI) method demonstrated accurate estimation and fast computational speed. However, the original MAGI method is limited to ordinary differential equations (ODEs), which are inadequate for some dynamic systems, calling for more complex or flexible structures in the specification of the DE model. Motivated by this, this thesis extends the framework of MAGI to facilitate inference for three common but challenging contexts, including (i) delayed differential equations, where system components exhibit time delays in their responses, (ii) mixed-effects ODEs, where experimental data consist of time-course observations on multiple subjects from a population, and (iii) selection of the most appropriate ODE model from a set of candidate models, where there is no true underlying model. The complex structures of these DEs introduce inferential and computational burdens and we address them in this thesis, along with computational and theoretical justifications. We illustrate the efficacy of our methodologies through simulated and real-world applications.
Content-Aware Pixel Art Rendering on Pixels of Multiple Shapes
(University of Waterloo, 2026-06-23) Wang, Zane Z.
Pixel art is a well studied art form that arose from technical limitations on computing hardware in the early 1980s. Although the discipline itself is often associated with video games, standalone character and landscape portraits in the pixel art style are also popular. Characterized by a deliberately limited resolution and colour palette, pixel art is as an artistic exercise in the conveyance of visual information with a limited number of samples, while avoiding certain unpleasant visual artifacts. In this thesis, we present a first solution to a novel problem in computer graphics: how do we render images in the pixel art style on other tilings of the plane besides the usual squares, all while respecting image features? We formulate the non-square (or "any-shape") pixel art rendering task as an energy minimization problem over tile-shaped filter supports, given a conventional raster image and geometric tiling data as input. We compute tile energy gradients via rasterization of the tiling geometry; using this information, we evolve an optimal filter support shape while imposing geometric constraints to balance between distortion and feature clarity. We then demonstrate that our method produces images with superior qualitative and quantitative properties in comparison with naive methods. Our program can compute finished images in seconds, and allows the user to watch the pixel art evolve in real time. We also provide some basic stylization and interaction features for artists, such as k-means colour quantization, colour palette generation in a perceptually uniform colour space, and brush-based vertex manipulation to adjust the shapes of the filter supports. This method has the potential to be useful in several artistic contexts, such as the creation of highly stylized portraiture and landscapes, and authoring of image and video for real hardware displays that use non-square pixels.
Exploring Inuvialuit youth food security experiences and supports in the Inuvialuit Settlement Region
(University of Waterloo, 2026-06-23) Ramirez Prieto, Maria
Background: The Inuvialuit Settlement Region (ISR) is in the Northwest Territories, Canada, and is the westernmost of the four Canadian Inuit regions. The ISR covers 906,430 km2 and includes six communities: Aklavik, Inuvik, Paulatuk, Sachs Harbour, Tuktoyaktuk, and Ulukhaktok. Today, Inuvialuit in the ISR are closely connected to and dependent on-the-land for physical, mental, spiritual, and emotional nourishment, which supports food security and well-being. Yet Inuvialuit youth face numerous barriers to participating in subsistence harvesting, and a growing body of literature documents a dietary shift away from country foods (CF) among younger generations. This is especially concerning as 69% of individuals aged 15 years or older in the ISR experience food insecurity. However, few studies have examined how Inuvialuit, including Inuvialuit youth, engage with CF and their relationships within the food web in relation to food security and well-being. While youth are the focus of the overall dissertation, Elders and families are also participants in this dissertation. Objectives: The purpose of this dissertation is to explore the web of relationships and experiences that shape the participation of Inuvialuit youth in the CF system, including the connection to the natural world, culture, and others, in order to generate community-driven evidence that fills gaps in the literature. Moreover, this dissertation aims to centre Inuvialuit knowledge holders through the co-production of knowledge, and use community based participatory action research (CBPAR) to conduct equitable and community-driven research. Methods: Using CBPAR, the studies in this dissertation employ diverse qualitative methods to conduct research with community members, including Community Research Leads (CRLs). Using a CBPAR approach provides an opportunity for research to move away from research on Indigenous communities to research with and for Indigenous communities and aligns itself with the National Inuit Strategy on Research. Photovoice, talking circles, and semi-structured interviews were used with purposive and snowball sampling of 11 Inuvialuit youth across all six ISR communities, 19 Elders in Aklavik, Paulatuk, Tuktoyaktuk, and Ulukhaktok, and nine families in Aklavik, Tuktoyaktuk and Ulukhaktok. Reflexive thematic analysis, using co-analysis methods, was used for all three studies. Results: Study 1 (Chapter 2) employed photovoice methodology, working with 11 youth participants who captured photographs of their CF experiences and shared ~5 photographs during semi-structured interviews. Through reflexive thematic analysis, our research team co-created five themes from the data: 1) CF supports Inuvialuit youth well-being; 2) preference for CF despite varied consumption and activity frequencies; 3) network of CF within communities; 4) strong foundational cultural knowledge and skills; and 5) cultural continuity. Study 2 (Chapter 3) brought together 10 youth from Study 1 and 19 Elders through talking circles to explore the relationship between youth, Elders, and intergenerational Inuvialuit knowledge (IK) transmission in relation to CF, food security, and well-being. In addition to semi-structured questions, photo-elicitation was used to initiate conversation between Elders and youth about the photograph’s subject matter and to invite storytelling (e.g., caribou harvest, goose roast for dinner). Our research team co-created four themes from the data: 1) fostering cultural connection and knowledge transmission through CF and family time; 2) emphasizing oral teachings as essential for well‑being; 3) recognizing the true cost of store‑bought food and goods; and 4) working together for community food security In Studies 1 and 2, family was identified as a crucial aspect of youth connection to CF and IK, in turn, supporting food security and well-being. As such, in Study 3 (Chapter 4), nine families (n = 28 participants) from Aklavik, Tuktoyaktuk, and Ulukhaktok were interviewed through semi-structured group interviews to explore the role of CF and family in the transmission of IK to support youth food security and well-being in the ISR. Our research team co-created four themes from the data: 1) learning on-the-land through experiences; 2) nourished by the land; 3) navigating barriers; and 4) the guiding principles for present and future generations’ well-being. Conclusion: Together, these studies examine Inuvialuit youth, Elders, and families’ experiences in the CF system, including identifying facilitators and barriers to accessing CF and IK. These studies make substantive contributions to the literature by documenting what Inuvialuit have long known – that CF is essential for youth, family, and community food security and well-being. Concurrently, these studies offer critical qualitative evidence that broadens the predominantly quantitative and store-bought-food-centered literature in the ISR. Adding to a growing body of literature, this research highlights that CF, along with the relationships it fosters with people, the land, culture, and community, supports food security while also nourishing the mind, body, and soul. This research employed a CBPAR approach, engaging community members at all stages of the research process and aligning with Inuit Tapiriit Kanatami’s National Inuit Strategy on Research and the Inuvialuit Regional Corporation’s ISR Research Data Strategy to ensure that Inuvialuit are included and respected as knowledge holders, thereby fostering respectful and beneficial research for Inuvialuit communities.
Effects of Noise on Optimization, Statistics, and Simulation of Quantum Systems
(University of Waterloo, 2026-06-23) Duschenes, Matthew
Understanding interactions between a system and its environment has consistently been at the centre of scientific studies. Indeed, environmental effects are far reaching and rich in behaviour, from heat baths leading to heat engines in thermodynamics, to non-conservative forces leading to dissipation in classical mechanics, to many-body interactions leading to emergent phenomena in statistical mechanics. As we transition from technologies based on classical phenomena, to technologies based on quantum phenomena, understanding system-environment interactions is the central challenge within quantum information sciences. These interactions arise frequently in quantum settings, due to intentional non-unitary dynamics as part of quantum algorithms, inherent random dynamics within chaotic systems, or unintentional environmental noise in the absence of error mitigation. It is thus essential to derive any underlying structure from these interactions, and to determine their implications on the viability of emerging quantum technologies. In this thesis, we conduct systematic analytical and numerical analyses of non-unitary dynamics. First, we study the practical effects of noise and experimental constraints on variational quantum algorithms. We find that objectives are initially robust to noise, and decrease exponentially with increased evolution time, before increasing polynomially with evolution time, due to an accumulation of errors. Second, we develop analytical tools to exactly compute statistics of ensembles of random quantum channels. Such formalisms allow us to derive hierarchies between ensembles, to define channel t-designs, and to show that generalized channel-design-induced concentration phenomena can occur. Third, we study distributions of probabilities of generalized measurement outcomes, given simulated noisy random quantum circuits. We develop an accurate and interpretable effective global noise model for these locally noisy distributions. Notably, we show that non-symmetric measurement distributions are multi-modal, whereas symmetric measurement distributions are uni-modal. Fourth, we propose and benchmark a classical simulation method, where measurement probabilities of states are represented by stochastic tensor networks, and non-unitary dynamics are represented by non-negative matrix factorizations. We conclude this thesis with a discussion of implications of the rich structures underlying non-unitary dynamics. We first interpret and provide examples and counter-examples of the utility of ensembles within channel-centric quantum algorithms. We proceed to discuss long term objectives posed by our studies, regarding constructing phase diagrams of optimization success, across ansatz expressiveness, noise scales, and system sizes. We also propose less passive applications of noise towards steering ensembles towards concentrated or non-concentrated behaviours. Finally, we raise questions of simulability of multi-modal distributions in the search for quantum versus classical advantage.
Scaling and generalization in neural quantum states
(University of Waterloo, 2026-06-23) Moss, Megan Schuyler
Because of their relevance to our understanding of quantum materials, the ground states of quantum many-body systems are a central object of study in condensed matter physics. However, the exponential growth of the Hilbert space with physical system size makes it difficult to study these states. Even with sophisticated numerical methods, we are often limited to studying finite-size systems that may not be representative of the thermodynamic limit, where properties of the system coincide with measurements of real materials. Scaling studies are therefore crucial for bridging the gap between what we can compute and what we want to understand. Quantum simulators, which are engineered and programmable quantum systems, offer an alternative approach to understanding quantum many-body systems. While these devices enable the direct preparation of certain quantum states, the need to verify the states prepared on these devices presents another exponentially difficult problem. The tools of modern deep learning, such as neural networks, have proven to be extraordinarily capable of extracting patterns in complex and high-dimensional data. Crucially, the learned patterns often correctly describe new data that the network was not exposed to during training, a phenomenon known as generalization. Despite belonging to exponentially large Hilbert spaces, quantum many-body ground states are often highly structured. Neural networks, which generalize precisely by learning and exploiting such structure, offer a promising approach to the problems of representing and characterizing such states. In this thesis, I focus on the use of neural networks to study the ground states of quantum many-body systems. When used in this context, neural networks are referred to as neural quantum states (NQS). Not only are NQS flexible and expressive ansätze, but they can be trained in different ways, depending on the information about the target state that is available. On the one hand, NQS can be trained with a data-driven approach, using measurement data from quantum simulators. We show that, because of their generalization abilities, NQS are poised to maximize the value of limited and imperfect data from experiments on today's quantum devices. On the other hand, NQS can be trained with a Hamiltonian-driven approach, which only requires knowledge of a system's Hamiltonian. Using this approach, we demonstrate how the generalization abilities of certain NQS architectures can be leveraged to enable efficient and accurate large-scale simulations of quantum many-body systems. Finally, we directly investigate generalization in the context of NQS, connecting our results to important observations in the broader machine learning research community. Together, these results demonstrate that the generalization abilities of NQS are not only essential for, but fundamentally linked to, their capacity to enable scalable studies of quantum many-body systems.
On Deep Learning for Nonprehensile Manipulation
(University of Waterloo, 2026-06-23) Caro, Steven
Nonprehensile manipulation, i.e. interaction without grasping, is a fundamental capability for mobile robots operating in unstructured environments, yet it remains a challenging control problem due to complex contact dynamics and under-actuated physics. Progress in this domain has been further hindered by the lack of standardized evaluation frameworks, leading to fragmented research efforts. This thesis addresses these gaps through two primary contributions: a unified benchmarking suite and a novel hierarchical control architecture. First, we introduce Bench-Push, a comprehensive benchmark designed specifically for pushing-based mobile robot tasks. Unlike existing benchmarks that penalize environment interaction, Bench-Push provides diverse environments — ranging from navigation-centric mazes to manipulation-centric delivery tasks — and introduces novel metrics to quantify the trade-off between task efficiency and interaction effort. We validate the framework by evaluating state-of-the-art baselines and demonstrating successful zero-shot transfer to a physical robot. Second, we propose the Hierarchical Reinforcement Learning - Diffusion Policy (HeRD), a hybrid architecture designed to solve long-horizon manipulation tasks. We identify that while Reinforcement Learning (RL) excels at strategic decision-making, it struggles to learn precise low-level contact dynamics. Conversely, Generative Diffusion Models synthesize smooth, context-aware trajectories, but lack high-level planning capabilities. HeRD bridges this gap by decoupling the control hierarchy: a high-level RL planner, utilizing a Spatial Action Map action space, selects strategic subgoals, which are then executed by a low-level, goal-conditioned diffusion policy. Extensive experiments in the Box-Delivery task demonstrate that HeRD significantly outperforms both state-of-the-art learning-based methods (SAM) and classical motion planners (Greedy Heuristic, Hierarchical RRT*). HeRD achieves higher success rates and greater interaction efficiency in both simulation and real-world deployments. Furthermore, we demonstrate that HeRD is capable of robust zero-shot generalization to unseen, unstructured clutter, successfully navigating complex environments where classical planners experience catastrophic failure.
Modernizing Informed Consent for Ubiquitous Personal Health Information Collection: A Blockchain Dynamic Informed Consent Platform for Public Health Research
(University of Waterloo, 2026-06-23) Da Silva E Souza Miranda, Pedro Augusto
Public health research is increasingly dependent on ubiquitous collections of personal health information (PHI) and personal data drawn from wearable sensors, electronic records, and connected devices. Traditional informed-consent processes, often paper-based and one-time agreements, cannot meet the demands of real-time data flows, cross-institutional sharing, and evolving regulatory requirements. This thesis proposes and tests a novel approach to dynamic informed consent, grounded in data governance theory and implemented through tamper-evident ledger technologies. First, it develops an Inter-Organizational Data Governance (IODG) framework tailored for public health research. Using the Governance Analytical Framework, the thesis maps actors, social norms, and transaction types, and allocates decision rights and accountability to researchers, data custodians, regulators, and data subjects. The framework emphasizes personal data sovereignty and prescribes a phased implementation, with validation, endorsement, and machine-readable policies to make PHI sharing lawful, transparent, and auditable. Second, the thesis translates this framework into the Consentio platform (“I consent”, in Latin), a web-based dynamic consent platform. Governance clauses are decomposed into formal requirements and encoded as use cases, class and state diagrams; propositional logic invariants enforce legal states; Angular route guards and validators prevent unauthorized actions; and a permissioned ledger records immutable, cryptographically verifiable receipts for approvals, consent, and data-sharing events. Third, a mixed-methods user evaluation assesses the platform. Ten participants (five researchers and five data subjects) reported high usability (global PSSUQ mean = 1.82 on a 1–7 scale), moderate to high digital skills and strong satisfaction, yielding a favorable Net Promoter Score. A reflexive thematic analysis of interviews and open-ended feedback revealed eight themes: participants valued granular control and centralized dashboards but sought clearer explanations of revocation limits and ledger security; progressive disclosure balanced efficiency and comprehension; and researchers desired customization, improved notifications, and audit-ready reporting. The findings suggest that coupling principled governance, formal system design and ledger-backed auditability can modernize informed consent management for data-intensive public health research, enhancing trust, autonomy and regulatory compliance. Future work should extend the framework to more diverse populations, refine interface elements for error recovery and ledger explanation, and evaluate scalability and interoperability across varied research domains.
Individual Open-Ended Problem Solving and Creativity
(University of Waterloo, 2026-06-18) Doroshenko, Sofiia
Complex real-world problems are often ill-structured and open-ended, with no single correct solution and many possible ways to organize the available information. Despite the prevalence of such problems, most experimental research on problem solving has focused on well-defined tasks with predefined solutions, while studies of ill-structured problems have largely relied on qualitative observational methods. This thesis extends a controlled experimental framework for studying open-ended problem solving, originally developed and tested in a group context by Alattas (2023), to the level of the individual solver. Sixty participants each completed three categorization tasks in which they organized 16 randomly selected pictures into four categories of four pictures each. Task open-endedness was manipulated by varying the goal structure and participants' beliefs about the solution space across three conditions: an Expert condition, in which participants were told to find the single best solution as identified by a panel of experts; a Good condition, in which multiple acceptable solutions were described as available; and a Story condition, in which participants were asked to form narrative-based categories. All participants completed all three conditions with different stimulus sets, using a within-subjects design. Quantitative analyses examined five outcome domains: task difficulty, solution variability, path dependency, association strength, and search behavior. Open-endedness significantly increased solution variability and path dependency, reduced association strength, and led to more direct search paths with fewer direction reversals. However, the effect on task difficulty diverged from the group-level pattern: Expert condition produced the highest difficulty across behavioral measures, while Good and Story were comparable, rather than the graded Expert > Good > Story ordering observed in groups. Post-experiment interviews were analyzed for 30 participants using an inductive thematic approach. Three main themes emerged: interpreting the pictures before categorizing, strategies for building categories under constraint, and using narratives and creativity to reach a solution. These themes illuminate the processes behind the quantitative patterns, revealing how participants interpreted ambiguous stimuli, managed constraints, and sometimes used stories to connect the pictures across all three conditions. Together, the quantitative and qualitative findings show that the effects of open-endedness on solution variability, association strength, and search behavior observed by Alattas (2023) in groups extend to individual solvers, whereas the relationship between open-endedness and task difficulty shifts at the individual level. The study contributes an experimental method for studying individual open-ended problem solving, provides early evidence on how the shift from group to individual processing changes the relationship between open-endedness and difficulty, and raises questions about the relationship between open-endedness and creativity that invite further investigation.
Phases of matter in quantum information and error correction
(University of Waterloo, 2026-06-17) Negari, Amirreza
This thesis investigates phases of matter and phase transitions through the lens of quantum information, with an emphasis on phenomena not fully captured by conventional local observables or equilibrium order parameters. While the traditional framework of phase transitions relies on correlation functions and order parameters, entanglement and other information-theoretic quantities provide a broader language for characterizing both equilibrium and non-equilibrium many-body systems. A central perspective developed here is that such quantities furnish sharp diagnostics of phases and criticality, particularly in topological phases subjected to noise and measurement. First, we study how measurements on topological quantum states reshape entanglement structure and induce phase transitions. Focusing on the toric code, we show that measuring part of the system generates distinct entanglement phases in the remaining degrees of freedom, and that tuning the measurement protocol drives transitions between them. To analyze these phenomena, we develop analytical tools that track the entanglement structure of the post-measurement state and reveal a rich phase diagram. Next, we turn to topological codes in the presence of noise, where information-theoretic probes reveal forms of non-equilibrium criticality invisible to conventional observables. In this setting, we identify extended critical behavior in mixed states and show that conditional mutual information diagnoses transitions between distinct regimes of information retention and loss. Interpreted through quantum error correction, these transitions distinguish phases in which logical information is robustly preserved, only partially accessible, or completely lost. Building on this connection, we extend the mixed-state perspective from static codes to fault-tolerant dynamics by relating faulty syndrome-extraction circuits to the mixed-state structure of an associated higher-dimensional resource state. This leads to a decoder-independent diagnostic of fault tolerance based on the conditional mutual information of syndrome data across spacetime. The resulting spacetime Markov length diverges at the fault-tolerance threshold, providing an intrinsic information-theoretic characterization of the preservation and breakdown of logical information in noisy quantum circuits. Finally, we develop structural results for thermal and symmetry-constrained mixed states. We show that symmetry can obstruct the sudden death of entanglement in thermal states: for canonical ensembles and for Gibbs states subject to superselection rules, entanglement persists, and in broad settings remains nonzero at arbitrarily high temperatures. In fermionic systems, this identifies parity superselection as a generic mechanism protecting mixed-state entanglement and fermionic negativity. Complementing this perspective, we study extendibility as a tractable probe of entanglement structure in fermionic Gaussian states, showing that it admits an efficient characterization and provides practical criteria for mixed-state entanglement, including an extendibility transition in the disordered Kitaev chain. Taken together, these results support a unified picture in which information-theoretic quantities serve as fundamental diagnostics of phase transitions and criticality in both equilibrium and non-equilibrium quantum systems.
Chemical Looping Combustion with an Industrial Waste: Kinetic Modeling and Pilot-Scale Design using Red Mud
(University of Waterloo, 2026-06-17) Ronson, Dana
Chemical looping combustion (CLC) is an emerging carbon capture process that can produce a high-purity stream of CO2 without the energy-intensive separation that is associated with traditional carbon capture strategies. In the process, a solid metal oxygen carrier (OC) facilitates the splitting of the conventional combustion reaction into distinct oxidation and reduction subreactions such that the fuel and air atmospheres remain separate. CLC has yet to be implemented at industrial scale; hence, there is interest in further developing this emerging technology. One such area of development is the OC material, as the overall performance of a CLC system is crucially dependent on the performance of the OC. While synthetic OCs have been the dominant materials used for CLC development, they demand valuable materials. Thus, there has been a recent interest in utilizing lower-cost materials such as industrial wastes in CLC. The use of industrial waste OCs in CLC has gained recent attention as these materials demonstrate the potential to be a cost-effective alternative to synthetic OCs. A key limitation in the development of CLC with industrial waste OCs is the lack of modeling efforts on CLC systems with these materials. This work presents a dynamic multiscale packed bed reactor CLC model to investigate the performance of red mud, an industrial waste from the alumina refining industry, as an OC. Kinetics describing the oxidation reaction of red mud with oxygen as well as reduction reactions of red mud with CH4, CO, and H2 fuels were identified and validated using lab-scale experimental data. Sensitivity analyses were performed on kinetic parameters and reactor operating conditions, where the model exhibited reasonable predictions. The model developed in this work serves to advance the development of CLC by enabling simulation and model-based design methods for the packed bed reactor with a red mud OC. A proposed nominal pilot scale design exhibits moderate utilization of the red mud OC and high fuel conversion. By producing approximately 848.1 MJ of energy in a single cycle, this design demonstrates the potential for red mud to be an effective OC in large scale CLC. The red mud pilot-scale design was compared to a similar system from the literature that used a synthetic OC, and it was found that the red mud system produced less heat as a result of its low density leading to a smaller solids inventory. Nevertheless, red mud boasts lower material costs than traditional synthetic OCs. An economic optimization of the pilot scale design for separate reduction and oxidation stages revealed that it is crucial to consider the integration of both stages to determine an optimal design for a complete cycle of the CLC system (i.e., jointly considering how the performance of reduction impacts the economics of oxidation).
Order in the Open: Symmetries and Entanglement of Many-Body Mixed States
(University of Waterloo, 2026-06-17) Almeida Lessa, Leonardo
Real-world quantum systems are open and interact with their environments, requiring a statistical description via mixed states. This thesis investigates the interplay between global symmetries and quantum entanglement in open many-body systems, asking whether symmetries can robustly enforce long-range entanglement and correlation patterns, even under severe decoherence or high temperatures. In the first half, we extend quantum anomalies to mixed states and establish the anomaly-nonseparability correspondence: mixed states that are strongly symmetric --- where every state in the statistical ensemble possesses the same symmetry charge --- exhibit long-range multipartite entanglement. We show that the unique multipartite structure of this anomalous entanglement gives rise to entirely new phases of matter that are intrinsically mixed, i.e., lacking any pure state representative. Conversely, we demonstrate that strong-weak mixed anomalies, such as Lieb-Schultz-Mattis anomalies, imply long-range correlations without strictly requiring quantum entanglement. Broadening this correspondence to higher-form symmetries, we introduce a definition of mixed-state phases of matter that is insensitive to long-range classical correlations, thereby only capturing distinct patterns of long-range entanglement. We argue that strong symmetries and their anomalies are the defining features of such phases. In the second half, we shift focus to non-anomalous symmetries and show when they alone suffice to enforce entanglement. We investigate maximally mixed states invariant under on-site symmetries, which naturally emerge as steady states of generic quantum evolutions that preserve these symmetries strongly. We exactly calculate the values of several entanglement measures that are notoriously difficult to tackle analytically or numerically, such as the entanglement of formation and distillation. For continuous non-Abelian symmetries, we find high amounts of long-range entanglement, despite the states being maximally mixed within the symmetric subspace. Finally, we prove that the same strong symmetry conditions and superselection rules prevent the sudden death of entanglement at finite temperatures, even for Abelian symmetries. This explains previously observed behavior in canonical ensembles with Ising symmetry and in fermionic systems.
Manu with a Ball: Water Entry of Two Tandem Spheres
(University of Waterloo, 2026-06-16) Chan, Michelle
A popular diving maneuver known as the “manu bomb” has long been a hallmark of recreational water activities in New Zealand. This cannonball-like dive generates a large splash and produces a pronounced air cavity beneath the water surface. As the cavity collapses, it generates a loud noise and focuses the surrounding fluid into a vertex where a vertical jet, known as the Worthington jet, is formed. If a diver performs the maneuver while holding a ball (e.g., a football), the ball is propelled upward by the Worthington jet, which we refer to as the manu with a ball or “manu ball” for short. As interesting as this dive with a ball is to witness, there are no existing studies on this phenomenon yet. In this work, we study the mechanism of the manu ball and provide a theoretical framework for maximizing the height, and thus the “fun”, of launching the ball. We model the manu ball as the tandem water entry of two spheres: the bottom sphere representing the diver, and the top sphere representing the ball. Our theoretical model quantifies the rebound of the top sphere as a momentum transfer ratio, comparing the initial and final momentum of the top ball over the initial and final momentum of the bottom ball. This momentum transfer ratio is a function of the dimensionless h1 number, a number representing the distance from the top ball to the pinch-off point normalized by the size of the bottom ball, which can physically be interpreted as the spacing between the top and bottom ball at water entry. This momentum transfer ratio is also parameterized by key factors such as the mass ratio between the top and bottom balls and jet strength. Our model was then validated by experiments, where the two spheres were positioned at a set initial separation and released with prescribed time delays. The process of water entry and rebound of the balls was recorded using a high-speed camera. The bottom ball was varied across four sizes and a range of weights to achieve different types of water entry-induced cavities, including both quasi-static and deep seal. Our experimentally validated framework provides a quantitative basis for understand- ing and optimizing the manu ball. By modeling the system as a two-sphere water-entry problem and identifying the governing non-dimensional parameters, we capture the essen- tial physics of jet formation and jet-ball coupling. The resulting scaling laws enable the prediction and enhancement of the top ball’s rebound, and establish a foundation for fu- ture investigations of recreational water-entry phenomena and related jet-driven propulsion mechanisms.

Browse

Recent Submissions