39456 research outputs found
Sort by
Robust Statistical Modeling In Functional Linear Regression
No full textFunctional linear regression is a prominent field within the domain of functional data analysis, with extensive applications in various domains such as biomedical studies, brain imaging, and chemometrics. However, despite the abundance of literature on functional linear regression, limited attention has been devoted to addressing outliers or heavy-tailed distributions in the data. Consequently, robust statistical analysis remains an underdeveloped practice in this area. The primary objective of this dissertation is to enhance the utilization of robust methods for modeling functional linear regression by primarily focusing on robust estimation techniques, hypothesis testing procedures that are resilient to outliers or heavy-tailed distributions, and robust variable selection methods.
First, we consider the problem of robust estimation in partial functional linear models under RKHS framework. The theoretical properties of robust estimation simulation studies are discussed in this chapter. Furthermore, two real data examples are presented to illustrate the performance of the robust procedure.
Then, we extend three robust tests: Wald-type, the likelihood ratio-type and F-type in functional linear models. Meanwhile, we investigate the theoretical properties of these robust testing procedures and assess the finite sample properties through the numerical simulation.
Finally, we propose a robust variable selection method in multiple functional linear regression and present a novel algorithm for identifying significant functional predictors using a robust group variable inflation factor (VIF) selection procedure. Our methodology is validated through rigorous simulation studies as well as its application to real-world data.
To ensure the cohesiveness of this dissertation, Chapter 1 provides an introduction to the research background, mathematical foundations, and primary motivations underlying this study. Chapter 2 presents a comprehensive overview of basis expansion methods for functional data analysis. Lastly, Chapter 6 concludes this dissertation by offering potential avenues for future research
Perceived Norms about Coping-Motivated Drinking Mediate the Relations Between Social Anxiety and Alcohol Use and Related Problems
No full textSocial anxiety has been shown to be a risk factor for alcohol-related problems among emerging adults. Perceived norms may be a relevant cognitive factor underlying this risk; however, previous literature has focused mostly on broad norms about alcohol use in general and not on norms about specific alcohol use behaviours. The main goal of this study was to examine the mediating role of perceived approval about drinking to cope specifically, in the pathway from social anxiety to alcohol use and related problems. I hypothesized that university students with heightened social anxiety would perceive their friends (in particular) as being approving of drinking to cope with negative affect, which would lead to an increase in alcohol problems. Participants were ages 18 to 29 from six Canadian universities and completed an online cross- sectional survey measuring social anxiety levels, perceived approval of specific risky drinking behaviours, alcohol use frequency, and alcohol-related problems. Mediation analyses showed that elevated social anxiety predicted greater perceptions of friends approving of drinking to cope, which in turn, predicted elevated alcohol-related problems. This mediation effect was not seen when analyzing perceived approval from typical students, highlighting a specificity for friends increasing perceived approval of coping-related drinking behaviours. Perceived approval of sexual risk taking from friends was associated with lower alcohol-related problems, and perceived approval of heavy drinking from typical students was associated with increased alcohol outcomes. This study is the first to examine the impact of perceived approval of specific risky drinking behaviours on alcohol outcomes among students with relatively elevated social anxiety. Such research could contribute to improving the efficacy of personalized normative feedback interventions for modifying normative perceptions and alcohol outcomes
Nonlinear Dynamics, Stochastic Methods, And Predictive Modelling For Infectious Disease: Application To Public Health And Epidemic Forecasting
No full textStatistical models must adapt to the evolving nature of many processes over time. This thesis introduces flexible models and statistical methods designed to infer data-generating processes that vary temporally. The primary objective is to develop frameworks for efficient estimation and prediction of both univariate and multivariate time series data. The models considered are general dynamic predictive models with parameters that change over time, featuring time-varying regression coefficients or variance components. These models are capable of accommodating time-dependent covariates and can handle situations where information is incomplete.
Several novel enhancements to existing mathematical models are introduced, with a particular focus on online learning and real-time prediction. Efficient Bayesian inference methodology is developed for analyzing the posterior of covariance components of dynamic models sequentially with a closed-form estimation algorithm for real-time online processing. Additionally, an online change detection algorithm for structural breaks is developed, combining the benefits of Kalman filters with sequential Monte Carlo methods.
A general and extensible compartmental model for the study of infectious disease data is proposed, with several innovative extensions to established probability models for the analysis of data. Next, we extend the classical SIRS (Susceptible-Infectious-Recovered-Susceptible) model by integrating innovative stochastic mean-reverting transmission processes to more accurately capture the variability observed in real-world epidemic data. Lastly, we provide a methodology that harnesses expansive data sources and feature engineering for analyzing and forecasting peak time and height of epidemic waves, crucial for the planning of public health strategies and interventions.
The performance of these inference methodologies is assessed through simulation experiments and real data from clinical, social-demographic, and epidemic domains
Resident Space Object Light Curve Classification & Space Situational Awareness Sensitivity and Simulation Studies
No full textThe number of objects being launched into space is rapidly increasing, emphasizing the critical importance of detecting, characterizing, and tracking these objects—an area of focus known as Space Situational Awareness (SSA). These Resident Space Objects (RSOs) include satellites (both active and inactive), rocket bodies and debris. Knowing the type of object near our satellites of interest is very important as it gives satellite operators the knowledge needed to accurately plan maneuvers to keep our orbits safe. This dissertation explores three main contributions within the field of SSA.
The first is a light curve classifier which uses Machine Learning (ML) to classify Low Earth Orbit (LEO) RSOs into stable satellites, tumbling satellites and rocket bodies. Multiple approaches were tested but the method with the highest accuracy is a Barlow Twins network which has a 75% accuracy for two minute light curves and a 97% accuracy for five minute light curves. The classification is used to characterize the motion of objects, which operators can use in combination with real images to determine the risk of collision and to perform effective maneuvers.
The second contribution is regarding SSA mission planning. A sensitivity analysis was conducted to determine the best camera to use for observing co-orbiting RSOs within 250 km of the observer. The analysis includes exploring the location of potential targets in the Field-Of-View (FOV) of the observer as well as the Signal-to-Noise Ratio (SNR) of different targets. A similar analysis to the one presented in this dissertation has been performed for the Redwing microsatellite mission.
Lastly, RSO image prediction simulations are tested for use in SSA. This dissertation demonstrated the implementation of an anti-sun pointing orientation for prediction simulations with validation from real images. Predicted images were used to determine targets for observation which were then validated following the downlink of the images
Transnational Adoptees' Identity, Belonging, And The Role Of Sport
No full textThis study examines how, and if, sport could influence transnational adoptees’ identity formation and sense of belonging. Transnational adoptees’ identity and belonging development is complex, dynamic, and contextual. For example, some transnational national adoptees are racialized individuals within a white society and feel a lack of belonging to either group. Three facets of identity were identified: pre-adoptive identity, the adoptee identity, and ethnic identity. Each identity facet was found to develop uniquely; a singular plot can not be used to understand the development of all the facets. Sport was found to offer transnationally adoptees a space to develop essential network connections, provide opportunities for ethnic identity exploration and cultural connection, and provide a reprieve from conflicts and tensions. However, pre-existing dynamics and social exclusion were also reproduced. The stories that emerged offer insights into how the context of sport influences transnational adoptees’ navigation of belonging and identity
Underwater gesture-based human-to-robot communication
No full textUnderwater human to robot interaction presents significant challenges due to the harsh environment, including reduced visibility from suspended particulate matter and high attenuation of light and electromagnetic waves generally. Divers have developed an application-specific gesture language that has proven effective for diver-to-diver communication underwater. Given the wide acceptance of this language for underwater communication, it would seem an appropriate mechanism for diver to robot communication as well.
Effective gesture recognition systems must address several challenges. Designing a gesture language involves balancing expressiveness and system complexity. Detection techniques range from traditional computer vision methods, suitable for small gesture sets, to neural networks for larger sets requiring extensive training data. Accurate gesture detection must handle noise and distinguish between repeated gestures and single gestures held for longer durations. Reliable communication also necessitates a feedback mechanism to allow users to correct miscommunications. Such systems must also deal with the need to recognize individual gesture tokens and their sequences, a problem that is hampered by the lack of large-scale labelled datasets of individual tokens and gesture sequences. Here these problems are addressed through weakly supervised learning and a sim2real approach that reduces by several orders of magnitude the effort required in obtaining the necessary labelled dataset.
This work addresses this communication task by (i) developing a traditional diver and diver part recognition system (SCUBANetV1+), (ii) using this recognition within a weak supervision approach to train SCUBANetV2, a diver hand gesture recognition system, (iii) SCUBANetV2 recognizes individual gestures, and provides input to the Sim2Real trained SCUBALang LSTM network which translates temporal gesture sequences into phrases. This neural network pipeline effectively recognizes diver hand gestures in video data, demonstrating success in structured sequences. Each of the individual network components are evaluated independently, and the entire pipeline evaluated formally using imagery obtained in both the open ocean and in pool environments.
As a final evaluation, the resulting system is deployed within a feedback structure and evaluated using a custom unmanned unwatered vehicle. Although this work concentrates on underwater gesture-based communication, the technology and learning process introduced here can be deployed in other environments for which application-specific gesture languages exist
Assembling The Digital Girl/girl: Making Meaning Through Social Media
No full textThis dissertation explores the digital becoming of girls through the various ways in which they are (re)made on social media. By using the thoughts and experiences of real girls, we explore together how we “make the Girl/girl mean” in our collective North American culture and society. Through interviews and focus groups with 23 girls located in the Greater Toronto Area, I developed six themes that outlines how the Girl (hegemonic discourses) is currently defined. Throughout my exploration of these themes, I critically analyze these definitions through an extensive review of girlhood, feminist, and social science literature. I bring into conversation previous research and theory with the emerging knowledge produced by the girls in this study and myself. My research creates a context specific roadmap, or what we might call an “assemblage” of the Girl as she exists in this current moment. Further, I think through how these Girl/girlhood subjectivities and discourses work in service to oppressive systems. I then think critically about what the Girl means to the girls in my study and real girls in general. In thinking through the lived, material realities of girls, I offer recommendations that can help us chart paths for the future, in which girls can be supported to safely exist in and explore this one wild and precious life
Exploring The Evolution Process And Requirements For Molecular Recognition Via Highly Functionalized Aptamers
No full textExpanding the chemical diversity of oligonucleotide libraries has allowed the evolution of synthetic nucleic acid polymers with enhanced molecular recognition and catalytic abilities. Thus, synthetic methods that enable the sequence-defined incorporation of diverse chemical modifications in developing novel or improved nucleic acid polymers and selection methods that facilitate efficient enrichment of high-quality aptamers are of particular importance in identifying novel or improved nucleic acid polymers for diagnostics and therapeutics. In this thesis, the advancement of ligase-catalyzed oligonucleotide polymerization (LOOPER) is discussed as a method to increase the chemical diversity of oligonucleotide libraries and its application towards the evolution of modified aptamers. An evaluation of the use of different ligases, scope and number of modifications, sequence space, and evolutionary outcomes from in vitro selections is provided, along with a critical lens on challenges to be addressed for the method to mature into a more widely adapted technology. Further to this, a variety of selection methods are discussed striving towards efficient single-round aptamer selection. The successful single-round aptamer selection against Thrombin is an inspiring outcome for future selections
The General's Son
No full textThe General’s Son is a 20-minute narrative drama that explores the raw, emotional complexity of General Cuba Wusalanga’s struggle with guilt, loss, and his longing for redemption. Set against a backdrop of allegorical African cultural elements, the film uses intimate cinematography, opulent aesthetics contrasted with isolating vastness, and symbolic imagery to represent underlying themes. The narrative underscores the tension and subsequent resignation of the General, as he grapples with the revelation of his monstrous nature. The vision intends for the audience to empathize with the General's predicament, inciting introspection and dialogue about societal cycles of corruption, and the potential to break free from them
Finding Common Ground: Methods For Sustaining Citizen Science Engagement That Increase Indigenous Plant Biodiversity In Southwestern Ontario
No full textNature-based solutions to address biodiversity loss require supports that reach beyond government grants. The Conservation Impact Bond (CIB) developed by Southwestern Ontario regional conservation charity, Carolinian Canada Coalition, is an example of a novel financial tool to incentivize biodiversity conservation by supporting citizen science.
My research evaluated Carolinian Canada’s In The Zone Tracker and its allied programs. A systematic review of the literature about Canadian citizen science projects provided context.
(1) The ITZ Program reversed plant biodiversity loss at a local level through planting projects that generated a self-reported increase in native species.
(2) Information about citizen science projects was difficult to discover. Academic research into citizen science projects published in peer-reviewed literature creates a more permanent record than web-based, grey literature.
(3) Citizen science projects do not necessarily improve science literacy. Rather, the ITZ tracker helped people to find common values and make positive, evidence-informed differences in their communities