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 17401

Semantic Ink: Intent-Based Graphics Authoring through Annotation-Driven Generative User Interfaces and Reusable Tools
(University of Waterloo, 2026-02-20) Li, Cara
Designers face increasing challenges when translating high-level creative intent into precise graphics operations. Traditional authoring tools expose low-level primitives that require complex sequences of tool invocations, imposing steep learning curves on users. While AI-assisted systems can interpret high-level intent, existing tools primarily support single-shot generation or style transfer, lacking scaffolding for iterative refinement and semantic-level reuse. We present Semantic Ink, an intent-based interactive graphics authoring system that combines freehand annotation with generative UI. Users express intent through annotations directly on the canvas, and the system responds by generating structured parameter spaces as interactive controls, for progressive refinement without prompt iteration. Successful edits can be extracted as custom tools for semantic-level reuse. Informed by a formative study with five designers, we contribute a two-dimensional taxonomy crossing five intent types (add, transform, style, remove, scene) with four controllable parameter types (variation, magnitude, spatialization, surface) to enable multimodal LLMs to parse ambiguous annotations into editable parameter spaces. A user study with six experts demonstrates that Semantic Ink improves editing efficiency, reduces cognitive load from mode-switching, and enhances personalization through tool reuse.
Iterated Function Systems and the Local Dimension of Measures
(University of Waterloo, 2026-02-20) Prandi, Joaquin Gabriel
Given an iterated function system $\mathcal{S}$ in $\mathbb{R}^d$, with full support and such that the rotation in it fixes the hypercube $[−1/2, 1/2]^d$, then S satisfies the weak separation condition if and only if it satisfies the generalized finite-type condition. With this in mind, we extend the notion of net intervals from $\mathbb{R}$ to $\mathbb{R}^d$. We also use net intervals to calculate the local dimension of a self-similar measure with the finite-type condition and full support. We study the local dimension of the convolution of two measures. We give conditions for bounding the local dimension of the convolution on the basis of the local dimension of one of them. Moreover, we give a formula for the local dimension of some special points in the support of the convolution. We study the local dimension of the addition of two measures. We give an exact formula for the lower local dimension of the addition based on the local dimension of the two added measures. We give an upper bound to the upper local dimension of the addition of two measures. We explore the special case where the two measures satisfy the convex additive finite-type condition that we introduce. We introduce the notion of graph iterated function system. We show that we can always associate a self-similar measure to the graph iterated function system.
Between Here and There: Architecture, Becoming, and the Search for Home After Displacement
(University of Waterloo, 2026-02-20) Correa, Samuel
This thesis investigates the persistent cycles of displacement and strategies of place-making among communities affected by forced migration, focusing on Colombia’s history of conflict as case study and situating its findings within a broader discourse on displacement. Drawing on social history, and analysis of theory, literature, and architecture, the work explores how the trauma of uprooting disrupts social ties, erodes a sense of belonging, and fragments cultural identity for those forced to leave their places of origin. At the heart of the thesis is the concept of “architectural design containers for counter-displacement offerings”, spatial interventions conceived not merely as shelters, but as active agents in fostering identity, resilience, and community. The proposed containers are analyzed through mixed methods, including, architectural precedent studies, cartographic analysis and site research, and theoretical synthesis, to understand how they may resist cultural erasure, support the rebuilding of networks, and nurture agency among displaced peoples. Central Northwest Toronto, an area shaped by both Colombian newcomers and a highly diverse immigrant population, becomes the setting for a network of architectural design proposals, chosen for its diverse urban forms and histories of migration. Through detailed site analysis and reference to international case studies, the thesis demonstrates how architectural design and intervention can meaningfully respond to the shifting needs of displaced communities, serving as a platform for reconnection, storytelling, and agency. By integrating theoretical frameworks, participatory research, and design praxis, the work asserts the important role architects and urbanists can play in reconciliation, restoration of dignity, and the creation of genuine belonging for those affected by displacement. The thesis advances the conceptualization of design containers as adaptable spatial frameworks that can be tailored to diverse displacement contexts, supporting resilience, healing, and the formation of lasting communities.
Use of Large Language Models (LLMs) in Qualitative Analysis: Evaluating LLMs as Assistive Coding Agents
(University of Waterloo, 2026-02-19) Neeb, Mikayla
Introduction: Large language models (LLMs) are increasingly used to support qualitative research, yet robust methods to evaluate the quality of LLM-generated codes remain underdeveloped. Existing approaches often rely on comparisons to human ground truth or custom evaluative methods, limiting cross-study comparisons. This study examines whether LLMs can function as assistive qualitative coding agents and introduces the CReDS framework as a structured approach to evaluating LLM-generated codes without the need for a comparative codebase. Methods: Two social media datasets were employed as validation sets to systematically develop and test approaches for evaluating LLM-generated inductive codes. Codes were generated using GPT-4o-mini and assessed through an iterative evaluation process. Initial assessment relied on conventional quantitative similarity metrics (e.g., cosine similarity); however, limitations in capturing qualitative distinctions prompted the incorporation of structured human evaluation. This process led to the development of the CReDS framework, comprising Consistency, Relevance, Distinction and Specificity, as a more comprehensive evaluative method. Targeted exploratory analyses further examined evaluative performance under specific conditions, further investigating the evaluative methods explored in this study. Results: LLM-generated codes aligned closely with human codes across both datasets, with overall semantic match rates ranging from 74-83%. At the text level, 65-95% of inputs had at least one LLM-generated code judged appropriate by human reviewers. CReDS scores revealed strong alignment to human-generated codes, with strong overlap across all dimensions. However, LLM-generated codes showed reduced specificity, and the CReDS framework observed conservative scoring behaviour. Despite these limitations, CReDS effectively surfaced systematic strengths and weaknesses in LLM outputs. Conclusions: These findings indicate that LLMs can reliably support early state qualitative coding when used as assistive tools under human oversight. The CReDS framework offers a transparent and scalable method for evaluating LLM-generated codes that align with qualitative principles while supporting iterative model development. This study contributes to a measurable and scalable platform for responsible human-AI collaboration in qualitative analysis and highlights directions for refining evaluation frameworks in future work.
A Child Friendly City: Redesigning Urban Spaces for Child Mobility and Play
(University of Waterloo, 2026-02-18) Villasmil Wilhelm, Sofia
Cities are rarely designed with children’s wants and needs in mind. Instead, they are shaped in ways that limit children’s opportunities for free play and independent mobility. These experiences are fundamental to children’s development and wellbeing, and their absence highlights a critical gap in contemporary urban design. This thesis investigates how such conditions shape and contribute to a child-friendly city and explores how urban environments can be redesigned to better support them. The research combines a literature review outlining the qualities that define a child friendly city, alongside an examination of the factors currently preventing cities from being considered child-friendly. It also includes a participatory workshop conducted with children to gain first-hand insight into their lived experiences, as well as a precedent analysis of places that are beginning to implement child-friendly interventions. Through this combined approach, the research identifies key spatial factors influencing children’s free play and independent mobility, including supported risk, flexibility, agency, and the inclusion of children’s voices. It also examines conditions and practices that should be avoided in child friendly urban design. These insights are translated into a set of adaptable design guidelines that prioritize children’s free play and independent mobility. Their application is demonstrated through three design proposals across sites of varying urban densities in Toronto, a city chosen for its wide range of urban conditions and openness to cultural and civic improvement projects. By positioning free play and independent mobility as central considerations in urban design, this thesis offers a practical framework for those seeking to create thriving and inclusive child-friendly cities.
Perceived Restorativeness and Restorative Outcomes: A Comparative Study of Diverse Environments in Urban and Natural Settings
(University of Waterloo, 2026-02-18) Grant, Emily
Natural environments are generally considered more restorative than urban ones, with various elements contributing to their restorative potential. This research investigates restoration in urban and natural environments using virtual reality and field studies. In particular, I examined urban and natural environments with both high and low restorative potential. Additionally, this research explores if participants can effectively evaluate an environment’s restorativeness based solely on the environment’s visual aspects. Across studies, restoration was assessed using subjective and objective measures of stress, attention, and affect. Results demonstrated that the natural environments did not consistently outperform the urban environments on restorative outcomes. Indeed, there were some indications that urban environments could also be restorative. Finally, participants’ predictions of restorative potential did not align with the restorative measures, indicating a gap between perceived restorative potential and actual restorative outcomes. Overall, the findings indicate that environmental restorativeness is complex, and not all urban or natural environments offer the same level of restoration. Further research is needed to understand the specific elements that contribute to an environment’s restorative potential.
Implications of a changing climate in coastal Labrador for caribou and their forage
(University of Waterloo, 2026-02-17) Lauriault, Patrick
The majority of Labrador’s coastal lands are below the 60th latitudinal parallel. Even so, the cold Labrador Sea currents, late-lasting sea ice, and frequent high winds, these coastal ecosystems often resemble Arctic and Subarctic ecosystems that are further north. Historically, these landscapes exhibited stunted shrub growth, but as the climate changes, the shrinking sea ice season and the cooling effects of coastal sea ice on nearby landscapes begin to subside, shrubs are now overtaking tundra vegetation communities. These shrubs threaten ground lichen communities in coastal regions. Understanding changes in vegetation allows us to predict changes in the forage available to caribou, a culturally and ecologically significant species in Labrador and other northern regions. Caribou tend to rely heavily on lichen in the wintertime to meet their dietary needs. In this dissertation, I addressed three main research questions: 1) How much lichen forage do caribou need to satisfy their energetic needs over winter? 2) With the state of ground lichen availability at three Labrador sites, what caribou density can be sustainably achieved at each location? 3) With climate change increasing coastal fog, how will lichen productivity respond to fog as a source of hydration? I used a time-series simulation to estimate caribou energetics over winter. After the simulation results, I assessed the state of ground lichen at three sites in coastal Labrador. The combined results are used to determine a sustainable caribou density at each location based on available winter forage. The caribou energetics simulation results showed that the average caribou must eat 1330 kg of lichen over the winter to avoid weight loss. The ground lichen estimates in open tundra at each of the three sites were 0.66 kg/m2 in Pinware (caribou free site), 0.1 kg/m2 in Cartwright (Mealy Mountain caribou) and 0.04 kg/m2 in Nain (George River caribou herd). With current lichen biomass estimates and an assumed 5% annual growth rate, I was able to derive sustainable caribou densities at each of the three sites (Pinware: 24 caribou/km², Cartwright: three caribou/km², Nain: one caribou/km²). I also studied how lichen productivity may be impacted by increased fog-based precipitation. Although lichens cannot compete vertically with shrubs, they may respond to climate change by becoming more productive when using fog water to increase growth in the growing season. Lichens readily use non-rainfall sources such as fog for their metabolism. Using a simulation model for lichen metabolism, I found that fog can encourage productivity in lichens, with my model showing a carbon uptake of 20 g on a 1 m2 ground lichen mat over four months from only observed fog events. Other promising findings from this study show that fog events happen much more frequently in the morning, hydrating the lichens before peak solar radiation. That fog alone will not block enough sunlight to achieve net photosynthesis. However, some troubling findings are that warmer months result in lower lichen productivity due to fog, as respiration begins to outpace photosynthesis at warmer temperatures. Fog water deposition is likely to increase in these environments, potentially altering lichen productivity in the north. Bottom-up constraints to caribou herds, such as a lack of forage, are essential to identify. Considering other threats caribou face can help herd managers and Indigenous people, who rely on caribou for a food supply, determine when intervention is required in a changing northern environment.
Fracture Characterization and Damage Accumulation Modelling of DP1180 Steel under Proportional and Non-Proportional Loading
(University of Waterloo, 2026-02-17) Jeyranpourkhameneh, Farinaz
Lightweighting remains a primary objective in the automotive industry, driven by the need to reduce fuel consumption and greenhouse gas emissions while meeting stringent crashworthiness standards. Advanced High Strength Steels (AHSS), such as Dual Phase 1180 (DP1180), have gained prominence due to their excellent strength-to-weight ratio. However, their complex fracture behavior under multiaxial and non-proportional loading conditions presents challenges for accurate failure prediction in structural simulations. This thesis aims to address these challenges through a systematic experimental investigation and modelling framework tailored to the fracture and damage response of DP1180 steel. The first phase of this work investigates the influence of gauge length on fracture strain in shear-dominated specimens. Conventional Digital Image Correlation (DIC) techniques were refined to enhance local strain measurement accuracy, focusing on strain localization in the shear zone. A series of tests were performed using butterfly shear specimens with varying gauge lengths to assess the lengthscale sensitivity of fracture strain. The results confirmed a strong dependence of measured fracture strain on the gauge geometry, reinforcing the need for standardized specimen design and DIC post-processing protocols. An optimized experimental configuration and robust DIC-based post-processing strategy were established to ensure consistent strain measurements for subsequent studies. The second component focuses on fracture under proportional loading conditions using uniaxial tension tests. Multiple specimen geometries were employed, including standard dogbone and notched samples, as well as conical hole expansion tests, to evaluate the fracture behavior of DP1180 under various constraints. Since fracture initiation under uniaxial tension is complicated by post-necking deformation, post-mortem surface strain analysis was performed to estimate local fracture strains. The study provided a reliable set of fracture strains for proportional loading conditions, allowing for direct comparison between different geometries and stress states. These results form the baseline for calibration and validation of fracture models under simple loading histories. The third phase of the work extends the investigation to combined loading paths involving simple shear and uniaxial tension. This approach enabled the evaluation of fracture behavior under intermediate stress states between pure shear and uniaxial tension. The resulting force-displacement responses and post-mortem strain measurements were used to validate the predictive capability of an existing phenomenological fracture model without necessitating re-calibration. The observed agreement between simulation and experiment under these combined stress states provides a robust validation of the model and highlights the versatility of the butterfly test methodology. To further extend the applicability of the framework, a novel experimental approach was developed to characterize fracture under non-proportional (bi-linear) loading paths. In this methodology, specimens were subjected to controlled proportional loading, after which miniature fracture specimens were extracted along different orientations and stress states. These samples were subsequently tested to failure, capturing the influence of pre-straining on fracture response. The collected data enabled an assessment of existing damage accumulation models under realistic forming conditions. Comparison with model predictions revealed that strain path changes significantly affect fracture strain evolution, especially for loading sequences that cross between tension- and shear-dominated states. These findings demonstrate the limitations of path-independent fracture criteria and underscore the importance of incorporating load history effects into damage modelling strategies. Overall, this thesis presents a comprehensive experimental framework for fracture characterization of AHSS under a wide range of loading conditions. The key contributions include: (1) development of a reliable shear fracture testing methodology that quantifies gauge-length sensitivity in DIC-based strain measurements, demonstrating variations in measured fracture strain depending on the selected length scale, (2) resolution of fracture strain identification under uniaxial tension through the combined use of multiple specimen geometries and post-mortem surface strain analysis, enabling the construction of a consistent proportional fracture dataset across a range of stress triaxialities, (3) validation of a phenomenological fracture model under combined shear–tension loading paths without re-calibration, showing good agreement between experimental observations and numerical predictions across intermediate stress states; and (4) development and application of a two-stage experimental methodology for evaluating fracture under non-proportional loading histories, providing a systematic assessment of path-dependent damage accumulation. Experimental results demonstrated that non-proportional loading generally leads to reduced fracture strains compared to monotonic proportional loading, with pronounced deviations governed by strain-path sequence and material anisotropy. Evaluation of the Generalized Incremental Stress State–Dependent Damage Model (GISSMO) showed that a damage exponent of 𝑛 = 2 provided the most consistent agreement with experimentally measured fracture strains across the investigated non-proportional loading conditions. Based on experimental repeatability and strain-field reliability, a hierarchy of confidence in the non-proportional fracture data was established, with v-bending tests exhibiting the highest confidence, followed by mini-biaxial, hole expansion, and shear tests. Collectively, these findings advance the understanding of path-dependent fracture and damage accumulation in DP1180 steel and provide experimentally validated guidance for improving the fidelity of forming and crashworthiness simulations involving advanced high-strength steels.
Tailoring Native and Transition Metal Catalytic Sites in Graphitic Carbon Nitride for Sustainable Material Design
(University of Waterloo, 2026-02-17) Pennings, Joel
Graphitic carbon nitride (g-C3N4) has emerged as a promising material for sustainable energy conversion applications due to its unique electronic structure, earth-abundant composition, and facile synthesis. However, its practical implementation is hindered by limitations in charge separation and catalytic activity. This thesis presents a comprehensive investigation into tailoring g-CN through advanced synthesis, exfoliation, and metal doping strategies, with a focus on enhancing its performance in air-metal batteries and photoelectrochemical systems. A novel femtosecond laser irradiation technique for g-C3N4 exfoliation is introduced, demonstrating superior control over layer thickness and defect density compared to conventional methods. The exfoliation process is optimized to yield exfoliated g-C3N4 nanosheets with tunable bandgaps and increased active surface area. Subsequently, a systematic study of metal doping (Cu, Fe, Ni, Co) on exfoliated g-C3N4 is conducted. The influence of dopant type, concentration, and incorporation method on the material's electronic structure and catalytic properties is elucidated through experimental characterization. Particular emphasis is placed on correlating the metal-nitrogen coordination environments with observed enhancements in charge transfer and oxygen reduction/evolution kinetics. The tailored M-g-C3N4 materials are then evaluated in air-metal battery and photoelectrochemical cell configurations. Viable improvements in battery capacity, cycle life, and conversion efficiency are demonstrated relative to undoped g-C3N4 and benchmark catalysts. Mechanistic insights into the enhanced performance are provided through in-situ spectroscopic studies and post-operation material characterization. Finally, the environmental impact and scalability of the developed materials and processes are assessed, providing a holistic perspective on their potential for real-world implementation in green catalysis and energy storage applications. This work establishes design principles for optimizing g-C3N4-based materials and demonstrates their promise as sustainable alternatives to precious metal catalysts in next-generation energy conversion devices.
Life Cycle Assessment of Milk Packaging in Canada: Evaluating Reusable Packaging under Prospective Energy Grid Scenarios
(University of Waterloo, 2026-02-17) Cha, Alexander
Plastic pollution remains a persistent environmental challenge in Canada, prompting growing policy attention toward single-use plastics. Yet, the life cycle implications of shifting materials or adopting reuse systems for food products, specifically milk packaging, are not well understood. This thesis conducts an ISO 14040/14044 aligned attributional life cycle assessment (LCA) comparing the environmental impacts of refrigerated milk packaging in Canada, including high-density polyethylene (HDPE) jugs, polyethylene terephthalate (PET) bottles, plastic milk pouches, liquid packaging board cartons, and glass bottles in both single-use and reusable systems. The functional unit is 12-L of pasteurized, refrigerated dairy milk contained in milk packaging. System boundaries are from cradle-to-grave for single-use formats and cradle-to-cradle for reuse scenarios, modeled in openLCA using the ecoinvent database and ImpactWorld+ impact methodology. Baseline packaging impacts are examined under current grid and recycling rates, and a scenario analysis is conducted under a prospective 2050 net-zero grid in Canada. Results show the packaging production stage dominates overall life cycle impacts, with lightweight flexible plastic milk bags performing best across most impact categories. Single-use glass exhibits the highest environmental impacts, while reusable glass shows improvement with increased reuse, though sanitization and transportation remain critical contributing stages as reuse increases. Under a decarbonized grid, reusable systems glass systems outperform select single-use packaging types but rarely surpass reusable plastic systems. Grid decarbonization improves electricity-intensive processes and strengthens relative performance of glass reuse, but does not outperform the advantages of lightweight, plastic-based systems. However, results are limited by data quality, availability, and quantification of microplastic impacts. Findings indicate that achieving circular packaging outcomes depends less on material substitution and more on system design. This study contributes novel, Canada-specific insight in how grid decarbonization and reuse interact in determining sustainable packaging strategy.
Analysis and Design of Lens Antennas for Power-Constrained Applications
(University of Waterloo, 2026-02-13) Esfarayeni, Nita
The ever-increasing demand for faster data rates as well as overcrowding in the sub-6 GHz spectrum has driven the shift to using higher frequency bands. While the use of higher frequencies can facilitate bandwidth requirements needed to meet the required data rates, they struggle with high Free-Space Path Loss (FSPL) which require specialized solutions to overcome. Phased Array Antennas (PAAs) have attracted immense attention in recent years. PAAs are able to make use of a large quantity of antennas to produce a gain high enough to overcome FSPL while also benefiting from compactness and the ability to rapidly steer and shape the beam. While they have been shown to be effective solutions for many applications, most PAAs depend on a large number of active amplifiers, which entails higher upfront costs, high power consumption, and high thermal dissipation. Such challenges must be addressed for power-constrained or heat-sensitive applications. This thesis presents a detailed analysis of existing solutions in literature and examines their trade-offs. The Lens Antenna Subarray (LAS) architecture is proposed as a solution, which offers low power consumption while keeping the Gain Over Noise Temperature (G/T) figure of merit for performance competitive with active arrays by leveraging both the directive properties of dielectric lenses as well as the flexibility of traditional PAAs. This thesis focuses on the design of a single lens and its feed network, referred to as a subarray. To produce a practical example, a Satellite Communication (SATCOM) receiver is chosen as the target application, and a subarray is designed. An ultra-wideband lens with a novel permittivity profile is designed which can provide up to ±64° of -3 dB steering, an improvement over similar Printed Circuit Board (PCB) compatible designs which typically do not provide more than ±50° of steering. Additionally, feed antennas are designed to provide wideband operation over the SATCOM receiving frequency range of 17.7 to 21.2 GHz. A total of 19 antennas are arranged in a novel feeding arrangement which enables the use of circular polarization with the lens, which has not yet been shown in literature. As the concept of LASs are relatively new to literature, there are many potential directions in which the concept can be developed. Further improvements to the lens, the simplification of feeding antennas, and array-level design are all areas which can be investigated in detail.
Assessment of interannual variability of benthic macroinvertebrate assemblages in the Athabasca River
(University of Waterloo, 2026-02-13) Chan, Cherrie
Interannual variability refers to the year-to-year changes in ecological conditions, which can alter niche availability and influence aquatic biota. Benthic macroinvertebrate (BMI) assemblages in riverine ecosystems have been shown to exhibit interannual variability in response to yearly variations in environmental conditions, in particular, large-scale fluctuations involving climate and hydrology. However, knowledge of interannual variation among BMI assemblages in cold region rivers is limited, highlighting the need to better understand year-to-year changes in BMI assemblages and how these changes are associated with environmental variability. To reduce this knowledge gap, this study quantified interannual changes in BMI assemblages and identified potential associations between hydroclimatic variables and assemblage composition along a cold region river. Benthic macroinvertebrates were collected annually over a range of 6 – 11 years from three reference locations on the mainstem of the Athabasca River, Alberta, Canada. Interannual variability of BMI assemblages was quantified by measuring change in assemblage abundance and taxonomic turnover, along with calculating change values in biological metrics between consecutive years. BMI interannual variability was then related to hydroclimatic variables (climate oscillations, temperature, and discharge) to determine potential environmental drivers. Results showed that interannual variability of BMI assemblages along the Athabasca River was similar to other studies conducted in cold region streams, and that four taxonomic groups (i.e., Chironomidae, Ephemeroptera, Plecoptera, and Trichoptera) were predominately responsible for yearly changes at all locations. The trajectory of interannual variability and the strength of environmental associations were inconsistent between locations. These results suggest that larger datasets are more likely to provide more reliable quantification of natural variability of BMI assemblages, and other unmeasured factors may influence interannual variability, thus confounding relationships with large-scale environmental drivers. Future studies should consider local-scale variability and finer-resolution environmental predictors to improve our understanding of the drivers of interannual variability.
Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference
(University of Waterloo, 2026-02-09) Basta, Petar
While it is known that quantization of large language models (LLMs) reduces memory usage via lower-bit weights, studies quantifying the resulting impact on energy usage and carbon intensity are scarce. We present the first unified evaluation of weight-only quantization strategies for LLM inference across varying input/output lengths and GPUs, highlighting differences in energy efficiency in addition to accuracy degradation and runtime. We study quantized inference on Llama 2 7B, Phi 3.5 Mini Instruct, OLMo 1B, Qwen 2.5 7B, and Qwen 2.5 14B across GLUE MNLI, MMLU, HumanEval, and GSM8K datasets, evaluating 10 post-training quantization (PTQ) strategies on NVIDIA H100, A6000, A100, and L4 GPUs. We identify a novel trend showing that quantization techniques tend to exhibit peak energy efficiency relative to full-precision baselines when inputs are sufficiently long and outputs are short. Furthermore, we show that quantization strategies become marginally more energy-efficient relative to full-precision models as batch sizes increase, though gains are modest. Notably, fused-kernel implementations such as EETQ int8 and Bitsandbytes int8 offer the highest energy savings, up to 4× compared to FP32 on short text generation, with negligible accuracy loss. Finally, we observe that energy usage closely tracks runtime on our evaluated benchmarks, indicating that, in practice, latency optimization can serve as a reliable proxy for reducing the environmental footprint of LLM services. We conclude with suggested directions for strategically selecting low-carbon quantization strategies based on specific inference requirements.
Lower Bounds on Average-Case and Non-Local Quantum Computation
(University of Waterloo, 2026-02-06) Asadi, Vahid Reza
This thesis studies computational and information-theoretic limitations in both classical and quantum models of computation. It is organized into two parts, each addressing a different aspect of computational hardness. Despite their differences, both parts share a common goal: understanding how structural and physical constraints shape what classical and quantum algorithms can achieve. In Part I, we present lower bounds on the average-case complexity of certain tasks in classical and quantum settings. We develop a general framework for constructing efficient worst-case to average-case reductions. Applying this framework, we obtain such reductions for fundamental problems in a variety of computational models; namely, algorithms for matrix multiplication, streaming algorithms for the online matrix-vector multiplication problem, and static data structures for all linear problems, as well as the multivariate polynomial evaluation problem. We further extend this framework to the setting of quantum algorithms, along the way obtaining a tight bound on the average-case quantum query complexity of the matrix-vector multiplication problem. Our techniques rely crucially on tools from additive combinatorics. In particular, we show local correction lemmas that rely on new probabilistic and noise-robust versions of the quasi-polynomial Bogolyubov-Ruzsa lemma. In Part II, we give quantum gate and entanglement lower bounds on certain non-local tasks. A non-local quantum computation (NLQC) replaces direct interaction between two quantum systems with a single simultaneous round of communication and shared entanglement. We study two classes of NLQC, f-routing and f-BB84, which are of relevance to classical information theoretic cryptography and quantum position-verification. We give the first non-trivial lower bounds on entanglement in both settings, under the assumption of perfect correctness. Within this setting, we give a lower bound on the Schmidt rank of any entangled state that completes these tasks for a given function f(x,y) in terms of the rank of a matrix G whose entries are zero when f(x,y)=0 and strictly positive otherwise. This also leads to a lower bound in terms of the non-deterministic quantum communication complexity of f. Due to a relationship between f-routing and the conditional disclosure of secrets (CDS) primitive studied in information theoretic cryptography, we obtain a new technique for lower bounding the randomness complexity of CDS. Finally, we show that the number of quantum gates plus single-qubit measurements required to implement a function f is at least linear in the entanglement-assisted simultaneous-message-passing communication complexity of f. As a consequence, we derive a linear lower bound for the inner-product function.
Ancestry Deconvolution via Differential Privacy
(University of Waterloo, 2026-02-05) Chowdhury, Raiyan
This thesis presents the first study of ancestry determination under differential privacy (DP). Direct-to-consumer genomics companies, such as 23andMe, offer ancestry testing to millions of individuals, yet remain vulnerable to severe data breaches. Such incidents are especially concerning because genomic data is uniquely identifying, highly correlated, and permanent once exposed. At the time of writing, 23andMe disclosed a catastrophic breach in October 2023 that compromised the genetic profiles of an estimated 6.9 million users, underscoring the urgent need for stronger privacy guarantees in genomic analysis. In this work, we investigate the application of DP to ancestry deconvolution. Using the 1000 Genomes dataset and Gnomix, a state-of-the-art ancestry inference model, we evaluate how privatizing single nucleotide polymorphism (SNP) data affects ancestry classification accuracy. We implement both naïve and correlation-aware local differential privacy (LDP) mechanisms across varying privacy budgets, enabling a systematic study of the privacy-utility trade-off in ancestry inference. Our results demonstrate that while naïve DP perturbations significantly degrade accuracy, correlation-aware LDP mechanisms preserve substantially more predictive power by accounting for linkage disequilibrium (LD). This thesis establishes a foundation for private ancestry deconvolution, providing an empirical benchmark of state-of-the-art DP methods in genomics and highlighting both the challenges and potential of integrating DP into ancestry testing.
NAVIGATING SETTLEMENT AND WELL-BEING: EXPERIENCES OF NEWCOMER WEST AFRICAN IMMIGRANTS IN ONTARIO
(University of Waterloo, 2026-02-04) Oko-Oboh, Ehinomen
Background: Canada has become a popular destination for immigrants, with a significant proportion arriving through economic immigration pathways. Despite the growing population, African economic immigrants remain understudied and are often grouped with other immigrant subgroups whose experiences may differ significantly from theirs. Although many arrive with strong educational and professional qualifications, African newcomers frequently encounter structural barriers and cultural transitions that shape their well-being, settlement experiences, and help-seeking behaviours. Understanding these experiences is important in improving culturally centered support. Objective: This study examined how settlement stressors, cultural perspectives, and changes in sociocultural realities shape West African economic immigrants’ settlement experiences, understanding of well-being, and their engagement with support services in Ontario. Methodology: This study employed a qualitative research design using semi-structured, in-depth interviews with nine West African economic immigrants who had lived in Ontario for between nine months and four years. Data were analyzed sequentially using Braun and Clarke’s six-step thematic analysis. Interpretation was conducted inductively and informed by Ager and Strang’s Integration Framework and Berry’s Acculturative Stress Theory. Results: Inductive coding identified six themes that captured the participants’ settlement journeys, identity negotiations, and well-being experiences. These themes were organized into three overarching domains: (1) Starting Over, which described participants’ lives before immigrating, emotional transitions, and early settlement stressors; (2) Identity and Systems Navigation, which highlighted cultural adjustments, racialization, experiences of ‘sudden’ Blackness, employment barriers, and housing challenges; and (3) Resilience and Well-being, which reflected key coping strategies, including faith, community support, and cultural maintenance. Conclusion: The findings highlighted that well-being among West African economic immigrants is multidimensional and shaped by the interaction of cultural identity, settlement challenges, racialization, and structural inequities. Overall, greater attention is needed to center the experiences of economic immigrants to develop meaningful and culturally responsive approaches to settlement support and integration.
Parent Anxiety and Child Psychopathology: A Longitudinal Study of Children with Physical Illness
(University of Waterloo, 2026-02-03) Elgie, Melissa
Estimates suggest that one in four children live with a chronic physical illness (CPI), such as arthritis, asthma, and diabetes. Children with CPI are vulnerable to developing psychopathology, known as multimorbidity. Approximately 40% of children with CPI experience multimorbidity, which is associated with lower self-esteem, poorer health-related quality of life, and a greater risk of substance use and suicidality. These negative outcomes may be a consequence of the substantial illness-related stress and uncertainty experienced by children with CPI and their families. To mitigate the compounding effects of child multimorbidity, it is essential to prioritize the health of these children by providing comprehensive, evidence-based healthcare services. However, doing so requires a deeper understanding of child multimorbidity, particularly in the family context. Beyond the affected child, parents of these children are likely to experience greater anxiety, resulting from managing CPI-related practical and emotional strains. This added stress and anxiety experienced by caregiving parents is posited to strengthen the association between parent and child mental health. Indeed, parent psychopathology has been identified as a key determinant of child multimorbidity; thus, efforts to improve the mental health of parents, in addition to the child, are paramount. Although there is consensus in the literature regarding the risk of children with CPI to developing multimorbidity, several knowledge gaps remain. Studies on child multimorbidity are predominantly illness specific, cross-sectional, and have short follow-up periods. Moreover, few studies have examined anxiety trajectories among parents of children with CPI, and the intersection between parent anxiety and child psychopathology. Although a handful of studies in non-clinical samples have explored associations between parent anxiety and child psychopathology, this knowledge is unlikely to generalize to the unique circumstances of childhood CPI. Accordingly, the aim of this dissertation was to obtain a more robust understanding of child multimorbidity, parent anxiety, and the complex association between parent anxiety and child psychopathology in this population. Opportunities to improve child and parent mental health via tailored, integrated, and family-centered approaches are emphasised. This dissertation is comprised of three distinct longitudinal studies which examined the onset of child multimorbidity, trajectories of parent anxiety, and the association between parent anxiety and psychopathology among children with CPI. Specific objectives were to: 1) evaluate child multimorbidity onset; 2) identify multimorbidity risk factors; 3) delineate trajectories of parent anxiety; 4) identify predictors of parent anxiety trajectories; 5) quantify the association between parent anxiety and child psychopathology; and 6) assess parenting stress and family functioning as moderators of this association. Data for this dissertation come from the Multimorbidity in Children and Youth across the Life-course (MY LIFE) study, a prospective cohort study of 263 children with CPI and their parents, who were followed for 48 months. The first study utilized survival analysis to determine that over the 48-month follow-up, 64% of children experienced multimorbidity (i.e., internalizing or externalizing psychopathology). Internalizing psychopathology was associated with greater child disability, older child age, and younger parent age, while the emergence of externalizing psychopathology was associated with male child sex and greater parent psychopathology symptoms. These findings suggest that a large proportion of children with CPI will develop multimorbidity and illustrates that multimorbidity onset is nuanced. Specifically, internalizing and externalizing psychopathologies were associated with distinct profiles of children. Findings highlight the importance of integrated physical and mental healthcare services to support the mental health of children with CPI. The second study examined the 48-month trajectories and predictors of parent anxiety symptoms using latent class growth modelling. Four trajectories of parent anxiety were identified: minimal, mild, moderate, and high. Approximately 40% of parents had persistent, moderate or high anxiety. Risk factors for less favourable anxiety trajectories were greater depression symptoms, higher educational attainment, having a female child, and having a child with multimorbidity. Results suggested that parents of children with CPI are at-risk of persistent anxiety, and parent- and child-related variables were associated with poorer anxiety trajectories. Because parents are responsible for the care of their children, promoting parental mental health is essential to ensure their well-being and that of their children. These findings garner support for greater refinement of pediatric healthcare services to be more family-centered and include strategies to promote the mental health of parents caring for children with CPI. The third study used linear mixed-effects modelling to examine associations between parent anxiety and child psychopathology. Results indicated that greater parenting anxiety was associated with greater child internalizing and externalizing symptoms over 48-months. Product-term interactions suggested that parenting stress and family functioning moderated the association between parent anxiety and child internalizing symptoms, while only family functioning moderated the association with externalizing symptoms. Results demonstrate the link between parent anxiety and child multimorbidity, and underscore the importance of evaluating parent anxiety, stress, and family functioning to support child mental health. These findings illustrate that to effectively manage the health of children with CPI, implementing evidence-based healthcare services which are family-centered must be prioritized. Knowledge from this dissertation addresses critical gaps in the understanding of child CPI, including multimorbidity onset, parent anxiety, and the association between parent anxiety and child psychopathology over time. Findings from this research reinforce that children with CPI and their parents face ongoing risk of psychopathology. Moreover, parent anxiety is associated with child multimorbidity, and family environmental factors likely influence this association. This research highlights five salient opportunities to refine pediatric healthcare: 1) early and routine mental health surveillance among children with CPI; 2) integrated physical and mental health services; 3) prioritizing family-centered care strategies; 4) targeted screening and interventions for at-risk subpopulations of children and parents; and 5) evidence-based interventions tailored to the unique experiences of children with CPI. Additional longitudinal studies are needed to explicate causal mechanisms which underpin the association between parent anxiety and child psychopathology, particularly among these vulnerable families. Providing comprehensive and integrated healthcare will help to reduce the incidence of multimorbidity and ensure the best possible outcomes for children with CPI and their parents.
Real-Time Speed of Sound Estimation for Point-of-Care Tissue Health Assessment.
(University of Waterloo, 2026-02-02) De la Torre Sanchez, Pat
Speed-of-Sound (SoS) is a fundamental acoustic property that emerging Ultrasound (US) modalities aim to leverage for tissue health assessment and image quality improvement. Tissue SoS has been tied to tumor malignancy classification, muscle health assessment, steatotic liver classification and bone porosity measurement among other. Consequently, leveraging the tissue SoS for more accurate beamforming, not only enables higher resolution imaging, leading to more accurate cyst classification, but also correct for heavy skull aberration and defocusing during High Intensity Focused Ultrasound. Directly measuring the tissue SoS requires multi-site access, thus limiting such methods to the breast and some limbs. Current Pulse-echo SoS estimation algorithms demand high computation time or provide a single global SoS value, either constraining real-time assessment or decreasing accuracy. In this dissertation, I developed: 1) A single-shot SoS estimation algorithm that estimates the global SoS of the media by leveraging the signal consistency across channels from a single transmitting event. 2) Multiple localized SoS estimators that leverage image dissimilarity from multiple transmission events to provide the user with the SoS of the segmented regions in the image, either in stratified media, or arbitrary configurations. By developing custom image formation, segmentation, raytracing and wavefront tracking frameworks, optimal transmission schemes, and GPU acceleration on a portable, research scanner, I’m able to provide robust, accurate, SoS estimation platforms that have been thoroughly validated in vitro, ex vivo and in vivo. This dissertation research aims to bridge the gap between SoS research and other US modalities that can benefit from the SoS information, as well as the gap between the low-level-research side that develops methods for robust US tissue assessment, and the clinical research side that tracks and relates such features to conditions of clinical relevance. With the single-shot algorithm, I provide non-SoS researchers with valuable information to increase the accuracy of their algorithms without the need of long transmitting times or high computational demands, crucial for flow or real time imaging. With the other algorithms, I provide robust avenues for non-invasive tissue health assessment, as well as arming clinicians with access to intrinsic tissue features, that can be used for real time monitoring or for subsequent research that arise from the insights clinicians gain with this novel tool.
Psychosocial Outcomes of Youth with Chronic Physical Illness and Siblings
(University of Waterloo, 2026-02-02) Basque, Dominique
Youth with chronic physical illness (YwCPI) are more vulnerable to developing psychopathology than their healthy counterparts. Prospective studies show that psychopathology in youth remains stable over time and extends into adulthood. Ongoing symptoms, treatment, and worry about disease progression increase the risk for worsened mental health. Given that health status often impacts physical, mental, and psychosocial health outcomes, health-related quality of life (HRQL) has been a focus in health research. The functional limitations resulting from disability caused by chronic physical illness (CPI) disrupt school participation, peer support, and family relationships, contributing to lower HRQL compared to healthy peers. Existing research on CPI continuity and HRQL is constrained by infrequent assessments and disease-specific approaches, limiting understanding of psychosocial outcomes in YwCPI. CPI in youth also affects siblings. This is consistent with family systems theory, which posits that emotional functioning results from interactions between an individual and their family members, and the family unit and its contextual circumstances. Research finds that CPI negatively impacts sibling mental, physical, and psychosocial health. Siblings of YwCPI are also at higher risk for internalizing and externalizing symptoms compared to controls. However, studies on sibling mental health and HRQL are limited. Few studies use longitudinal methods, limiting the ability to assess the long-term impact of CPI. The aim of this research is to assess psychosocial outcomes in YwCPI and siblings. To address the aforementioned literature gaps, the objectives of this dissertation were to: 1) compare the homotypic and heterotypic continuity of psychopathology in YwCPI and siblings over 48 months, 2) model 48-month trajectories and assess predictors of HRQL in YwCPI and siblings, and 3) assess the mediating effect of sibling psychopathology on the association between YwCPI disability and sibling HRQL. Data come from the Multimorbidity in Children and Youth Across the Life Course Study; data were collected at baseline, 6, 12, 24, and 48 months. Study 1 utilized cross-lagged panel modelling to assess internalizing and externalizing symptom continuity in YwCPI and siblings. Auto-regressive pathways assessed homotypic continuity and cross-lagged pathways assessed heterotypic continuity. Separate models were computed for YwCPI and siblings. The method of variance estimates recovery (MOVER) assessed how patterns of continuity in YwCPI differed from siblings. Significant (p<0.05) autoregressive pathways were detected between all time points for internalizing and externalizing symptoms in YwCPI and siblings. Few cross-lagged paths were significant in either model. The MOVER showed no significant differences in the magnitude of any path estimates between YwCPI and siblings. Study 2 delineated trajectories of change for each domain of HRQL (physical well-being, psychological-being, autonomy and parent relations, peer and social support, school environment) using latent class growth analysis. Models were computed for YwCPI and siblings independently. Backward stepwise regression assessed predictors of trajectory group membership. The number of trajectory groups ranged from two to four, and class size ranged from 5% to 80%. Over half of the models included an increasing, no change, and decreasing group, with most changes occurring at earlier time points. The proportion of siblings having no change in HRQL was greater than YwCPI. Group membership predictors were age, income, and parenting stress. In Study 3, linear mixed-effect models assessed the mediating effect of sibling psychopathology on the association between YwCPI disability and sibling HRQL. YwCPI disability predicted increased sibling psychopathology, which was associated with lower sibling HRQL on all domains. Mediation effects were statistically significant in all models. Given the homotypic continuity of symptoms of psychopathology, health services should promote early and routine mental health screening for YwCPI and siblings. Early changes in YwCPI and sibling HRQL trajectories highlight critical opportunities for family-centred approaches to pediatric care. The mediating effect of sibling psychopathology in the association between YwCPI disability and sibling HRQL underscores the importance of integrating siblings into family-centred care, with targeted mental health screening and interventions that aim to mitigate negative psychosocial outcomes.
Navigating the Politics of Food System Futures: Post-Disaster Food Sovereignty in Puerto Rico
(University of Waterloo, 2026-02-02) Diaz, Ileana
Puerto Rico has been devastated by protracted economic collapse. Hurricanes Irma and Maria’s landfall in 2017 shattered the archipelago, followed by earthquakes in 2020, the COVID-19 pandemic, and Hurricane Fiona in 2022. There have been chronic issues with recovery post-disasters. In the context of economic, political, and ecological uncertainty, one of the most pressing concerns for Puerto Rico now and in the future will be its food supply. This research explores the effects of sequential disasters on the people of Puerto Rico and their food system. The study investigates food insecurity, government responses, and the will toward food sovereignty. The study is derived from individual in-depth interviews (n=84), archives (legal/court documents), media analysis (newspapers/photographs), and government data. The research adopts a critical and decolonial framework to the study of post-disaster Puerto Rico.

Browse

Recent Submissions