21090 research outputs found
Sort by
Robust Nonparametric Inference on Manifold Spaces
No full textWe propose rank-based procedures for robust and nonparametric statistical inference on manifold spaces. Particularly, we focus on the problems of multi-sample hypothesis testing, multiple change point analysis, and statistical process monitoring when data lie on a Riemannian manifold. These methodologies provide a unified framework to deal with various types of data structures such as matrices, curves, surfaces, networks, to name a few. These types of datasets frequently appear in a broad set of applications such as communication networks, manufacturing, computer vision, autonomous systems and robotics. We evaluate the proposed methods considering various types of object data such as matrices, curves, text mining data, networks, shape data and landmarks.
In Chapter 2, we develop robust and nonparametric methods for hypothesis testing when data lie on a manifold. We demonstrate that ranks generated from data depth can be used for two-sample and multiple sample hypothesis testing of change in location and scale parameters. Several important properties of these tests such as asymptotic convergence, size and power, robustness with respect to qualitative-robustness and breakdown point are developed under mild nonparametric assumptions. These tests have several advantages, they have a simple distribution under null, they are computationally cheap, and they enjoy invariance properties. We demonstrate the efficacy of these methods with a numerical simulation and a data analysis. We show that these tests are robust when data are heavy tailed or skewed, and have higher power compared to their competitors in some situations, while still maintaining a reasonable size.
In Chapter 3, we propose robust and nonparametric single and multiple change point detection methods for stochastic processes defined on manifolds. These methods consider a variant of CUSUM statistic which is based on the rank of data depth. We demonstrate that changes in the rank of depth values can be used to detect change in the distribution of data lie on manifolds. To detect more than one change point, we consider binary segmentation and wild binary segmentation algorithms along with the proposed data depth rank CUSUM statistic. We demonstrate that both of these algorithms are consistent estimators of the number of change point(s) and the location of change point(s). In addition to asymptotic results, we develop nonasymptotic sharp bounds for single and multiple change point estimators. These test statistics can be applied to both intrinsic and extrinsic
manifold analysis frameworks. In simulation, we compare our methods against several methods from the literature, and demonstrate that the proposed methods outperform their competitors in some situations where dataset is contaminated with outliers. We also present the application of our methods to vehicle health monitoring, traffic monitoring on highways, and mall pedestrian surveillance.
In Chapter 4, we extend these methods to the setting of statistical process monitoring. We investigate statistical process monitoring scheme on general metric spaces, and propose exponentially weighted moving average, CUSUM, and Mann-Whitney moving average Shewhart control charts using rank of data depth. These methods are nonparametric and robust to outliers through the use of data depth ranks. We show that when sample size is large, our methods have simple behaviour under the null hypothesis. Since our methods are based on data depth ranks, we do not need the estimate of covariance operator which makes our method computationally cheap. Such advantages make these methods a favorable choice for online process monitoring. We demonstrate the robustness of these methods theoretically and numerically. We extract several nonparametric control charts from the literature for comparative study. Simulation results indicated that the proposed methods
outperform their competitors in many situations in terms of out-of-control average run length, while keeping the in-control average run length at a reasonable level. We present the application of our methods to laser power-bed fusion additive manufacturing process.
In Chapter 5, we present some possible directions for future research related to dynamic network and longitudinal data analysis on Riemannian manifolds. It is anticipated that the contributions achieved in this thesis will be applicable to a wide range of interdisciplinary research problems
Performance Analysis of Zero-Forcing Beamforming Strategies for the Uplink and Downlink of MU-MIMO Systems with Multi-Antenna Users
No full textMulti-User Multiple-Input Multiple-Output (MU-MIMO) systems provide high performance through the transmission of multiple data streams to users (resp. from users) on the Downlink (DL) (resp. Uplink (UL)) at the same frequency and time. Multiple antennas at the users can further improve performance by enabling the choice between diversity (using the multiple antennas per user to create one data stream with potentially higher rates) and multiplexing (utilizing the antennas to create multiple independent data streams per user). To achieve the potential of MU-MIMO systems with Multi-Antenna (MA) users, the Radio Resource Management (RRM) processes must be carefully executed. In this work, we focus on the DL and UL of realistic MU-MIMO systems with MA users.
The necessary RRM processes are user selection, digital beamforming (called here precoding), Power Allocation (PA), Power Distribution (PD) and Modulation and Coding Scheme (MCS) selection. We consider the well-known family of precoding strategies called Zero-Forcing (ZF), which nullifies inter-stream interference. There are three primary ZF-based precoding strategies: Coordinated-Transmit-Receive-1 (CTR1), where we enable only the strongest stream per scheduled user, Block Diagonalization (BD), where all possible streams are enabled per selected user, and Coordinated-Transmit-Receive-Flexible (CTRF), which allows a flexible stream allocation per scheduled user. The latter has the potential for increased performance compared to the other strategies at the cost of higher complexity.
Throughout this work, we conduct novel offline (where run time is unimportant) studies to compute the performance of MU-MIMO systems under Proportional Fairness (PF) with those precoding strategies on the DL and UL in realistic systems characterized by 3rd Generation Partnership Project (3GPP)-based scenarios where the Base-Station (BS) has a large number of antennas and employs practical MCSs. To enable the offline study on the DL, we employ heuristics from the literature and adapt them to consider practical systems with MCSs and PF, whereas we propose the necessary tools for the offline UL study. The offline studies allow us to execute performance comparisons of MU-MIMO systems with the precoding strategies (BD, CTR1 and CTRF), which can guide the design of real-time RRM heuristics for both the DL and UL. The DL results indicate that CTRF outperforms the other strategies under PF. However, depending on the network parameters, either CTR1 or BD could replace CTRF given their comparable performance. On the UL, the conclusions are similar, but in the 3GPP Urban Macro scenario, CTR1 presents a comparable performance to CTRF over all considered network parameters, emerging as an alternative to CTRF
Analysis of Heterogeneities in a 20 L Bioreactor
No full textBiological systems are utilized in various industries to produce valuable products, including biopharmaceuticals. This is done in bioreactors, which are specialized vessels that are able to precisely control key parameters, including agitation, air flow, temperature, pH, dissolved oxygen, and nutrient supply. With the high demands for biopharmaceuticals caused by advancements in medicine, the need for efficient production and optimization of bioreactors has been evident. This has been especially seen during the COVID-19 pandemic, and the high costs of some products, which are inaccessible to many individuals. To optimize production, simulation models have been developed to predict effective control schemes for high growth and product yield. However, this is challenging to translate between lab-scale and industrial-scale due to the formation of gradients in industrial-scale systems, which have poor mixing. Gradients lower the efficiency of bioreactors as cells must constantly adapt to changing extracellular conditions, which cause stress and lower yields. Thus, it is necessary to validate simulation models using the gradients formed in large-scale bioreactors; however, this data is not readily available, and it is difficult to obtain such gradients in smaller-scale bioreactors.
In this work, fed-batch experiments are studied to investigate the formation of gradients in dissolved oxygen, kLa, pH, cell density, glucose, and acetate concentrations. This was done through the movement of sensors, turning the air on and off, and the usage of different sampling locations. The objectives of this work were first to characterize the culture with flask and batch experiments and then to use this information to carry out the fed-batch experiments to explore the potential of measuring these gradients. Dynamic metabolic responses were observed and measured depending on the control of the glucose feeding, and consistent gradients were observed for the dissolved oxygen, pH, and kLa, while gradients for cell density, glucose, and acetate were not observed, which may be due to limitations in sampling times.
Finally, the metabolic responses have been modeled using modified Monod kinetics, where the modifications include self-growth inhibition, an acetate metabolic switch, and a cell density-dependent lag function. This work was done using a genetic algorithm on Python to optimize parameters, and the model was able to adapt to the different extracellular conditions presented in the fed-batch experiments
Visual Standards in Aviation: Implications for Safety, Performance, and Training Assessment
No full textObjectives: The goal of this thesis was to provide evidence to either support the current aviation visual medical standards or to begin to define new ones. Two aspects of vision (visual acuity and contrast sensitivity) were intentionally degraded to assess their effects on flight performance. Four experiments were conducted that examined this topic. Experiment one assessed the flight performance of novice pilots during two simulated flight scenarios: approach to landing or a short flight circuit and landing during clear and calm weather conditions with vision degraded with scattering lenses. Experiment two built on this by having novice and intermediate pilots completed a simulated flight circuit and landing during three adverse weather conditions (high wind, heavy rain or high wind and heavy rain) with their vision degraded by scattering lenses or by defocusing lenses. Experiment three examined whether vision could be used as a probe to evaluate flight instructors. This study had pilot participants complete a simulated flight circuit and landing scenario with their vision degraded using both scattering lenses and defocusing lenses while two flight instructors subjectively assessed their flight performance. The fourth and final experiment expanded on these studies by assessing both the flight performance and stress responses of pilot participants during a simulated flight circuit and landing while their vision was degraded either by scattering lenses or defocusing lenses.
Methods: Twenty participants were recruited for study 1. Pilot participants completed either an approach to landing flight simulation or a short flight circuit and landing simulation in clear, calm weather conditions with their vision degraded with scattering lenses to either 6 or 8 levels of degradation respectively. Twenty-six pilot participants were recruited for study 2 where they completed a short flight circuit and landing simulated flight in three weather conditions (wind, rain, wind and rain) with 5 levels of vision degradation. Vision was degraded by either scattering lenses or defocusing lenses. Study three examined using vision as a probe to assess flight instructor agreement and repeatability. Twenty pilot participants completed a simulated short flight circuit with their vision degraded using both scattering and defocusing lenses to 8 different visual acuity levels. There was a total of five flight instructors recruited for study three in which two instructors were present to assess the performance of a single pilot participant. This was repeated on a second day with the same pilot participant and the same flight instructors. Study four examined the effect of vision degradation (either scatter or defocus) on both the flight performance and on pilot stress levels (which were monitored via eye-tracking and heart rate sensors) in thirty-seven pilot participants. These participants were tasked with completing a short flight circuit and landing and some of the participants had randomized minor (increased oil pressure) or major (engine failure) emergency scenarios introduced.
Results: Results from all four studies showed that flight performance (vertical speed, airspeed, altitude, pitch, roll, landing hardness, landing accuracy) was not significantly impacted by mild and moderate visual degradation. Only severe degradation had an impact on performance. These studies also show that pilot participant contrast sensitivity may be a better indicator of performance than visual acuity, as declines in performance with milder vision degradation when using defocusing lenses compared to scattering lenses were seen. Studies two through four highlighted the importance of using both objective and subjective grading when assessing pilot performance as it was identified that there was a difference in when objective flight performance metrics (significant at 6/18 with scattering lenses and 6/60 with defocusing lenses) were affected and when flight instructors perceived a change in performance (significant at 6/18 with scattering lenses and 6/120 with defocusing lenses). The final study showed that while at several of the vision degradation levels flight performance was unaffected, pilot participant heart rate and eye-tracking metrics were affected. With heart rate variability and eye tracking metrics (saccade amplitude and saccade velocity) being affected at a degradation level of 6/18.
Conclusions: These findings challenge the strict reliance on the use of just visual acuity in the current aviation medical standards. Across four experiments, results showed that flight performance remained largely unaffected when defocus lenses were used but was affected at a much lower visual acuity degradation when scattering lenses were used, suggesting contrast sensitivity may be a more relevant predictor of pilot performance. The studies suggest that current visual medical standards may be unnecessarily strict, thereby restricting the pool of eligible pilots. This work also highlighted the need for a more standardized and evidence-based approach to pilot evaluation. The final experiment further underscored the role of stress in flight safety, as physiological stress responses were observed even when flight performance remained stable. Taken together, these findings suggest that a more comprehensive approach – incorporating contrast sensitivity, objective performance measures, and stress responses – may lead to a more effective and fair pilot vision standard
Systems and Methods for Generating Large Arrays of Optical Traps in Neutral Atom Array Quantum Processors
No full textNeutral-atom array quantum processors provide a scalable and controllable platform for quantum simulation, computation, sensing, and metrology. Strong Rydberg interactions enable entanglement and high-fidelity gates, while the ability to rearrange atoms allows for multi-qubit operations between spatially distant qubits and the implementation of quantum error-correcting codes with nonlocal stabilizers. These arrays also support electromagnetic field sensing and form the basis for next-generation atomic clocks with improved stability and precision. Recent progress has demonstrated platforms with thousands of qubits, the implementation of error-correction codes, and experimental realizations of quantum spin models, including studies of quantum phase transitions. Current efforts now focus on scaling to even larger arrays and reducing quantum gate errors, with the goal of achieving computations and simulations beyond the reach of classical devices.
However, increasing the scale and controllability of neutral-atom array platforms to address problems where quantum advantage can have practical impact remains challenging. Key difficulties include scaling to larger numbers of qubits, implementing fault-tolerant computation, and performing gates between spatially distant qubits. A critical experimental challenge is ensuring that qubits form an indistinguishable spatiotemporal ensemble, so that atoms across the array and over multiple experimental shots can be treated as identical. Achieving this uniformity becomes more difficult as the array size increases, due to the added complexity of ensuring that all atoms experience consistent control parameters, such as trap depth and magnetic fields. Without sufficient spatiotemporal uniformity, it is impossible to meaningfully average measurements across different qubits and experimental runs, undermining scalability and limiting the performance of quantum simulations, computations, and sensing applications.
In this thesis, I propose, design, and implement methods to scale up the number of qubits and improve the spatiotemporal uniformity of qubit properties in acousto-optically generated trap arrays for neutral-atom quantum processors. These methods enable magnetic-field imaging across an array of 1,305 optical traps containing 690 qubits, as well as high-fidelity fluorescence imaging of single atoms in optical tweezers.
First, I demonstrate a novel method for circularizing and de-astigmatizing the trapping beam using three cylindrical lenses. This approach is cost-effective, power-efficient, and broadly applicable. By improving the beam shape, we reduce the per-trap laser power, enabling the use of larger qubit arrays in neutral-atom quantum processors. Applied to a Ti:sapphire laser with an initial circularity of 0.69, this method achieves a circularity of 0.97 and a beam waist separation of 0.8 percent of the Rayleigh range, reducing the optical power required per trap by 5 percent.
After beam circularization, I develop a real-time closed-loop feedback system for an optical trap array generated by two orthogonal acousto-optical deflectors to enhance the spatiotemporal uniformity of qubit properties. The system stabilizes trap depths using power measurements from a fast CMOS camera and in-situ depth estimates from an EMCCD camera based on the atomic signal. Noise is decomposed into uncorrelated temporal modes, each regulated by an independent PID loop. Unlike approaches that regulate only total laser power, our method stabilizes the depth of individual traps, eliminating the need for frequent recalibration of cooling-beam parameters and enabling higher duty cycles. In a 45 by 29 array, the generated traps exhibit a standard deviation of 6 percent relative to the mean trap depth, limited by the number of independent control parameters actuating the acousto-optical deflectors.
Having established these capabilities, we demonstrate the controllability and practical utility of the platform by performing magnetic-field sensing using microwave horn spectroscopy on a 45 by 29 trap array, mapping field gradients across the array. This establishes the foundation for advanced quantum sensing protocols, such as quantum lock-in amplification, and paves the way toward entanglement-based quantum sensors capable of offering a quantum advantage in sensing.
Lastly, to improve readout fidelity, I describe the optimization of imaging beam parameters in neutral-atom array quantum processors. This optimization maximizes atom classification accuracy while minimizing the probability of ejecting atoms from the traps. Applied to a 45 by 29 trap array, the method achieves a classification fidelity of 99.98 percent and a loss probability of 0.12 percent with a 75 millisecond imaging time. These improvements reduce measurement errors, enhance quantum state readout, and increase the sensitivity of neutral-atom-based quantum sensors
Selling Character: Trade Papers, Modernity, and The Maclean Publishing Company, 1887-1910s
No full textThis dissertation explores the gendered dimensions of the growth of trade papers and trade paper advertising in Canada from 1887 until the 1910s using the lens of the Maclean Publishing Company. In 1887 John Bayne Maclean established The Canadian Grocer. The paper advised retail grocers on how best to succeed in the growing, ever-competitive retail environment, but Maclean’s ultimate goal was to profit from the paper through the sale of advertising space. In the late 1880s and 1890s, Maclean would establish and acquire additional trade papers including Books and Notions, Hardware, The Canadian Dry Goods Review, and Printer and Publisher, creating Canada’s most successful trade paper publishing company of the period. Maclean faced many challenges to building his business, including an initial level of distrust surrounding trade papers themselves and manufacturers’ initial refusal to purchase advertising space. This thesis seeks to explore the ways in which Maclean sought to overcome these impediments
Employing the Maclean trade journals, the Maclean-Hunter archival records, and the John Bayne Maclean papers as source material, this dissertation demonstrates how notions of manhood that were popularized in success literature beginning in the 1860s were used to sell advertising space, build the company’s readership base, Maclean’s own reputation, and that of his trade papers. The first chapter establishes context, providing the history of the Canadian trade paper field. The second chapter is a discursive analysis of Maclean trade paper content, demonstrating the ways in which Maclean applied notions of character to build his reputation as a publisher and grow his readership base. The third chapter explores Maclean’s use of nascent public relations techniques to publicize company policies to articulate his desire to serve the success of the retailers, wholesalers, and manufacturers he served. The fourth chapter explores personal and textual advertising space sales tactics. Here Maclean argued that advertising was akin to masculine performance and provided potential advertisers with the directives needed to articulate who they were as men through their advertising practices. This dissertation contributes to the historiography on gender and business in Canada by underscoring the ways in which masculinity influenced the emergence of trade paper advertising, trade paper publishing, and consumer culture more broadly
Vision-Based Adaptive Impedance Control for Soft Material Manipulation
No full textThe manipulation of soft materials presents a longstanding challenge in robotics due to their nonlinear, variable, and often unpredictable physical properties. Traditional control strategies either require complex physical modeling or extensive data collection, limiting their applicability in dynamic, real-world settings. This thesis proposes a model-free, vision-based adaptive impedance control framework that enables collaborative robots to regulate contact forces during manipulation using only real-time visual feedback. By dynamically adjusting stiffness based on vision measurements of object deformation, the framework allows for compliant and adaptive interaction with a wide range of soft materials without prior knowledge of their physical characteristics.
The effectiveness of the framework is demonstrated through two representative studies. The first, VAIRO (Vision-based Adaptive Impedance-control Robotic framework), addresses the challenge of rolling croissant dough in a craft bakery setting. This study involved using a standalone stereo vision camera to monitor the rolling process of puff pastry into croissants with a robotic manipulator. Employing RGB-D data and real-time image processing, VAIRO detects and segments croissant layer thickness and inter-layer gaps, which are used to adapt the stiffness of a Cartesian impedance controller. The result of the study showed VAIRO creating croissants closer to artisinal quality compared to a constant stiffness approach, where VAIRO produced croissants with up to 6 times less variance in heights and mean error of heights less than 0.51mm.
The second study, VAISI (Vision-based Adaptive Impedance-control for Surgical Incisions), applies similar methodology to perform depth-controlled incisions in soft biological tissues using point-clouds captured by a stereo vision-enabled scalpel end-tool on the robotic manipulator. Using geometric measurements on the point cloud, the scalpel's depth and skin tissue deformation is tracked in real-time to adapt both the incision depth and force through visual servoing and stiffness adaptation of the Cartesian impedance controller. VAISI demonstrated the capability to create incisions with sub-millimeter accuracy and a maximum variance in incision depths of 1.23mm.
The vision-based adaptive impedance control framework presented in this thesis offers a foundation for future research in flexible automation, medical robotics, and other domains requiring interaction with soft materials. This contributes towards the goals of Industry 4.0—namely, adaptable and robust robotic systems
Advancing Proteomic Analyses with Graph-Based Deep Learning: Protein Inference and DIA De Novo Peptide Sequencing
No full textProteomic analysis plays a central role in unraveling the complex molecular underpinnings of biological systems. However, traditional approaches to protein inference and peptide sequencing have been hampered by challenges such as data complexity, label scarcity, and spectral noise. In this thesis, we leverage advanced deep learning techniques to address these challenges, thereby expanding the efficacy of proteomic analyses.
Our work is organized around three major contributions. First, we introduce GraphPI, a novel protein inference framework that redefines the inference problem as a node classification task within a tripartite graph structure. In GraphPI, proteins, peptides, and peptide-spectrum matches (PSMs) are modeled as interconnected nodes, while edges incorporate features such as peptide identification scores and a specialized peptide-sharing attribute. By harnessing a tailored graph neural network (GNN) architecture inspired by GraphSAGE, our approach effectively aggregates and propagates information across heterogeneous node types. Critically, GraphPI is trained in a semi-supervised manner using pseudo-labels generated from established protein inference methods, combined with hard negative decoy information. This training process not only circumvents the typical bottleneck of limited labeled data but also yields protein scores that generalize across diverse datasets, all while substantially reducing computational overhead relative to Bayesian network–based approaches. Experimental evaluations on multiple benchmark datasets demonstrate that GraphPI delivers competitive accuracy with significant speed improvements, thus paving the way for real-time applications in large-scale proteomic studies.
Second, we present DIANovo, an innovative deep learning method designed to tackle the inherent complexities of Data-Independent Acquisition (DIA) data for de novo peptide sequencing. Unlike conventional de novo approaches that often struggle with the multiplexed nature of DIA spectra, DIANovo incorporates a suite of strategies to manage coelution and spectral noise. Our approach begins by constructing a spectrum graph that captures the mass differences between peaks. Next, a Transformer-based encoder, enhanced with Rotary Positional Embeddings (RoPE), processes the graph by encoding these mass differences along its edges, effectively treating the spectrum graph as fully connected. Furthermore, DIANovo introduces a coelution-aware pretraining stage, where the model is first optimized to predict ion types from coeluting peptides. This pretraining step equips the network with a nuanced understanding of spectral interferences, thereby improving the fidelity of subsequent peptide sequence predictions. In addition, a two-stage decoding strategy is employed: the first stage identifies an optimal path through the spectrum graph, while the second refines this path to generate a final amino acid sequence by filling in mass tags. Comparative analyses against state-of-the-art methods reveal that DIANovo achieves significant improvements in both amino acid and peptide recall, especially when applied to high-quality narrow-window DIA data obtained from next-generation instruments such as the Orbitrap Astral. Moreover, we investigate whether DIA identifies more peptides than DDA in de novo sequencing by comparing their performance on the same biological sample under varying acquisition modes and parameters. Our results demonstrate that DIA only outperforms DDA when employing narrower isolation windows.
The third component of this thesis presents a comprehensive theoretical analysis that sheds light on the performance limits of peptide identification methods. By linking the signal-to-noise profile to peptide identification accuracy, our study elucidates the inherent trade-offs between Data-Dependent Acquisition (DDA) and DIA strategies. We derive quantitative metrics to predict peptide identification performance under a range of experimental conditions, and these predictions are validated against empirical data. This framework not only explains why Astral DIA data can provide superior peptide coverage in certain scenarios but also delineates the conditions under which peptide identification is most favorable. These insights are crucial for guiding the design of future mass spectrometry experiments and for optimizing computational pipelines in proteomic research.
Collectively, the three contributions of this thesis demonstrate the transformative potential of integrating deep learning with advanced computational frameworks in proteomics. GraphPI and DIANovo both showcase how novel neural network architectures can overcome longstanding challenges in protein inference and de novo peptide sequencing, while the theoretical analysis provides a foundation for understanding and further refining these methodologies. The experimental results across multiple datasets underscore the robustness, efficiency, and generalizability of our approaches, suggesting that deep learning–based strategies will play an increasingly central role in the future of proteomic analysis.
In conclusion, this work not only advances the state-of-the-art in protein and peptide identification but also offers practical solutions for handling large-scale, complex proteomic data. By bridging the gap between theoretical insights and practical implementations, our integrated framework lays the groundwork for enhanced biomarker discovery, more accurate disease diagnosis, and a deeper understanding of biological systems at the molecular level
Conjunctive Queries with Negations: Bridging Theory and Practice
No full textAntijoin, given its significant expressive power, has numerous applications in relational data analytics. Notwithstanding its importance, there remains great research potential in antijoin processing. In practical database systems, existing techniques to process antijoins are still considered rudimentary, building upon heuristics and cost-based optimization strategies that offer no theoretical guarantees. Meanwhile, the database theory community has proposed algorithms for antijoins with strong theoretical guarantees, yet these algorithms build upon specialized, complicated data structures and have not made their way to practice. In light of such gap between theory and practice, we propose new algorithms for antijoin processing in this thesis. Not only do our new algorithms provide the same theoretical guarantees as the state-of-the-art algorithm, but they also use only basic relational operations. The latter property enables our new algorithms to be rewritten in basic SQL statements, allowing an easy, system-agnostic integration into any SQL-based database system. We then empirically evaluate one of our new algorithms, rewritten in SQL, over real-life graph datasets with a variety of SQL database systems. Experimental results show order-of-magnitude improvements of our new algorithm over vanilla SQL queries
Resistance is Our Heritage: An Archive of Survival and Efforts to Resist Gentrification in Little Jamaica.
No full textRooted in my experience as a community member and organizer, this thesis documents the history and resistance of Little Jamaica in Toronto as it continues to face the impacts of gentrification and displacement. Using Black archival practice as my methodology, I draw from oral histories, protest materials, community reports, and digital media to center the voices and experiences of residents, business owners, and activists. Grounded in the framework of racial capitalism, this research understands gentrification as part of a longer history of displacement, extraction, and state neglect targeting Black communities. It traces the development of Little Jamaica through Caribbean migration and examines how planning interventions—particularly the Metrolinx Light Rail Transit construction—have intensified economic pressure, disrupted local business, and contributed to cultural erasure. By amplifying community narratives and mobilizing knowledge for advocacy, this thesis not only documents the fight to preserve Little Jamaica’s cultural identity, but also contributes to broader discussions on gentrification, displacement, and resistance. It highlights the everyday strategies, care networks, and collective organizing that continue to sustain the neighborhood. Ultimately, it seeks to equip residents with knowledge and tools for organizing while challenging dominant narratives of progress and revitalization. At its core, this work affirms that the fight for Little Jamaica is ongoing—and that resistance has always been part of its story