123129 research outputs found
Sort by
First-Principles Study on Periodic Pt2Fe Alloy Surface Models for Highly Efficient CO Poisoning Resistance
Full text linkSurface and sub-surface atomic configurations are critical for catalysis as they host the active sites governing electrochemical processes. This study employs density functional theory (DFT) calculations and Monte Carlo simulations combined with the cluster-expansion approach to investigate atom distribution and Pt segregation in Pt-Fe alloys across varying Pt/Fe ratios. Our simulations reveal a strong tendency for Pt atoms to segregate to the surface layer while Fe atoms enrich the sub-surface region. Crucially, the calculations predict the stability of a periodic Pt2Fe alloy surface model, characterized by specific defect structures, at low platinum content and low annealing temperatures. Electronic structure analysis indicates that forming this Pt2Fe surface alloy lowers the d-band center of Pt atoms, weakening CO adsorption and thereby enhancing resistance to CO poisoning. Although defect-induced strains can modulate the d-band center, crystal orbital Hamilton population (COHP) analysis confirms that such strains generally strengthen Pt-CO interactions. Therefore, the theoretical design of Pt2Fe alloy surfaces and controlling defect density are predicted to be effective strategies for enhancing catalyst resistance to CO poisoning. This work highlights the advantages of periodic Pt2Fe surface models for anti-CO poisoning and provides computational guidance for designing efficient Pt-based electrocatalysts
Word meaning extension in humans and machines
Full text linkWords in natural language often take on new meanings, so that language users can express an infinite set of intended meanings with a finite vocabulary. Such processes of word meaning extension (WME) are highly productive, diverse, but also non-arbitrary. However, it remains unclear what common cognitive mechanisms and knowledge are driving various types of word meaning extension. Modeling WME is also relevant for natural language processing (NLP), since novel lexical expressions are constantly emerging through time, and NLP systems should effectively interpret and generate such novel word usages in a human-like way. In this dissertation, I develop a computational framework that not only accounts for word meaning extension in historical language development, but also supports NLP systems to flexibly interpret and generate novel word usages.
My dissertation is organized into two main parts. In the first part, I study word meaning extension from a computational cognitive science perspective. I first propose a probabilistic generative model of word meaning extension that relies on multimodal semantic knowledge, and show that the model, when incorporated with the cognitive processes of chaining, can accurately predict historical emergence of novel verb-noun syntactic compositions. Next, I present a novel and general account of semantic chaining through the lens of cognitive efficiency. This account explains different chaining mechanisms as a tradeoff between representation and complexity. I show that the efficiency-based framework can be formulated as an infinite mixture model from Bayesian non-parametric statistics, which adaptively constructs word meaning through time under limited computational resources.
In the second part of my dissertation, I study how modeling human-like word meaning extension can enhance natural language processing. Specifically, I propose the problem of word sense extension, where a neural language model trained on limited linguistic data is asked to generate novel usages based on previously unseen word senses. I develop a generative framework that combines deep few-shot learning with semantic chaining to capture incremental word sense extension, and I show that this framework can be leveraged to fine-tune and improve word sense disambiguation on rare word senses. Furthermore, I investigate how modeling the systematicity in word meaning extension in combination with language models helps the construction of non-literal word usages like metaphors. I show that learning the analogical similarity between word meanings effectively improves language model systematicity in making both incremental and irregular types of word meaning extension. I therefore suggest that learning systematic word meaning extension benefits language models on multiple tasks pertaining to figurative language understanding.
In summary, my dissertation contributes a principled paradigm for modeling the generative processes of word meaning extension, and it opens up future opportunities for human-like automated processing of creative language use.Ph.D
Piggutuit Qanuqtuurniutuk; An Indigenous Flower, Resourceful in the Face of Challenges
Full text linkPiggutuit Qanuqtuurniutuk: An Indigenous Flower, Resourceful in the Face of Challenges is a botanical allegory that centers Indigenous knowledge systems, land-based relationships, and the ongoing impacts of colonialism. Framed through the metaphor of piggutuit—a resilient, native flower—the narrative explores how Indigenous plants (and peoples) have long thrived in reciprocal relationship with the land, only to be disrupted by the arrival of invasive species that seek dominance over harmony. This story captures the destructive influence of colonization through ecological metaphor, illuminating efforts to displace, control, and erase Indigenous presence, while also celebrating the enduring strength, adaptability, and resurgence of Indigenous lifeways.
Rooted in Inuit Qaujimajatuqangit and Indigenous relationality, this piece blends personal narrative, ecological insight, and academic literature to challenge dominant settler-colonial paradigms. It reflects on the deep wisdom embedded in Indigenous botanical relationships and proposes that decolonization requires collective responsibility, interspecies ethics, and the restoration of land-based relationships. The work invites botanists, educators, and knowledge-keepers to reconsider the distinction between flower and weed—not as a scientific fact, but as a reflection of cultural judgment. Ultimately, this submission is a call to action: to cultivate shared gardens where all beings—human and more-than-human—can thrive in balance, respect, and kinshipThe presentation of the authors' names and (or) special characters in the title of the pdf file of the accepted manuscript may differ slightly from what is displayed on the item page. The information in the pdf file of the accepted manuscript reflects the original submission by the author
Age, stable isotopes, and redescription of the first mammoth tooth from the Labrador Peninsula
Full text linkA mammoth (Mammuthus) molar was discovered on Long Island, Nunavut in 1878, making it the first mammoth fossil from the Labrador Peninsula and most northeastern mammoth in eastern North America. On the basis of morphology, we cautiously identify the tooth as the worn stump of the third left upper molar of a woolly mammoth (Mammuthus primigenius) although it was originally identified as the premolar of a Columbian mammoth (Mammuthus columbi). The Long Island tooth yielded an infinite calibrated radiocarbon age of >39 800 years. However, bulk-sampled stable oxygen isotopes in the tooth’s enamel structural carbonate suggest a local climate with a mean air temperature of around 2.1 °C. This roughly corresponds to the predicted temperatures in the Long Island area during climatic optima. The tooth’s age can further be constrained by the presence of glaciers in the Long Island area and the appearance of M. primigenius-type mammoths in North America. Thus, we estimate the Long Island tooth to be between 500 ka and 55 ka years old, but more likely from an interglacial period, and most likely from MIS5e. The Long Island tooth has a collagen δ15N value of 10.7‰, higher than expected for comparable North American mammoths and possibly indicative of nutritional stress. The tooth’s collagen δ13C value of -21.6‰ falls within the range of values for comparable mammoths, and the calculated dietary δ13C value of -26.0‰ suggests a diet rich in C3 plants.The presentation of the authors' names and (or) special characters in the title of the pdf file of the accepted manuscript may differ slightly from what is displayed on the item page. The information in the pdf file of the accepted manuscript reflects the original submission by the author
The Cultural Politics of Experimental Jazz Singing
Full text linkThis dissertation is a study of select historical and contemporary vocalists who perform experimental jazz singing. I examine both their vocal techniques and the politics of their performance. Exceedingly little research addresses experimental vocal techniques in music that is mapped by scholars, audiences, and/or practitioners as jazz in any of its many guises. Yet from Abbey Lincoln screaming on We Insist! (1960), to Jay Clayton ululating on Sound Songs (1986), and present-day musicians such as Fay Victor, DB Boyko, and Christine Duncan, singers have been contributing to practices that they and their audiences view as related to jazz, while also, in some cases, pushing the boundaries of what jazz “is” by stretching beyond aesthetic and technical conventions that dominate(d) the jazz scenes in which they were involved. I engage with the music of artists such as these, those whose voices and musical contributions continue to be largely overlooked by scholars, audiences, and pedagogues. My close analysis of the sounds, techniques, and contexts for these performers’ innovations and interventions is framed by a critique of value structures that have rendered them all but inaudible except in certain spaces. As such, this dissertation has the potential to help readers expand their sound palette and query some of the assumptions underpinning their perspectives on music, especially jazz, in the first place.
This work begins with a foundational chapter, which explores the lives and innovations of three Black American women all born in the 1930’s, who were central to the development of this music: Abbey Lincoln, Jeanne Lee, and June Tyson. In the following chapter, I move on to examine the career of vocalist Jay Clayton, who was crucial to the development of the Loft Scene in New York City in the 60s and 70s. Next, I examine the 2006 record Idiolalla by contemporary Canadian vocalists Christine Duncan and DB Boyko and percussionist Jean Martin. This chapter examines fears of vocal damage in this setting as well as recuperation techniques. Finally, this work ends with a compendium of 107 extended vocal techniques found throughout my research. This compendium includes descriptions as well as recordings of each technique.D.M.A
The Costa Rican Marimba Tradicional: Pedagogical Practices, Preservation, and National Identity
Full text linkWhile the marimba tradicional (traditional marimba) now holds an important place in Costa Rican culture, sustaining its practice remains an ongoing effort. Since 1970, the Costa Rican government has led preservation initiatives, prompting public universities to establish programs primarily focused on instructing musicians in the technicalities of playing the marimba tradicional (Acevedo Álvarez and Tablada Pizarro 2019; Umaña Picado 2022; Vida Estudiantil Universidad Nacional n.d.) and at rescuing its associated repertoire (Acevedo Álvarez and Guevara Duarte 2007; Jeremias 2008). Along with these efforts, Costa Rican scholars have conducted historical research on the marimba tradicional, its music, and its performers; however, questions about the current state of the practice remain understudied. In particular, there is a notable gap in research concerning the impact of these preservation initiatives, particularly their effect on the practice and teaching of the marimba tradicional and its cultural significance.
In response, this dissertation draws on fieldwork to learn about and critically assess the current state of the practice and teaching of the marimba tradicional in two settings: local communities (Guanacaste and San José) and institutional music programs (municipal music schools and public universities). The study examines how pedagogical methods, values, and performance practices differ and intersect across these settings. Furthermore, this analysis is situated within the framework of dominant notions of Costa Rican national identity, a discourse that not only informs governmental initiatives in cultural preservation but is also continuously influenced by the enduring impact of colonialism. The dissertation also explores how musical and pedagogical aspects of the marimba tradicional are shaped by broader societal concerns, revealing a complex cultural landscape where its practice and dissemination navigate the tension between its recognition as a national symbol and the persistent challenges of low cultural appreciation. Finally, this dissertation offers a critical understanding of concert marimbas, emphasizing the cultural and historical contexts often overlooked in Western music education. It challenges the detachment of the concert marimba from its Central American heritage, advocating for a deeper appreciation of oral pedagogies and the cultural traditions that shaped the identity of the marimba tradicional.
Aunque la marimba tradicional ocupa hoy un lugar importante en la cultura costarricense, mantener su práctica requiere esfuerzos constantes. Desde 1970, el gobierno de Costa Rica ha liderado iniciativas de preservación, impulsando a las universidades públicas a establecer programas enfocados principalmente en la formación técnica de músicos en la interpretación de la marimba tradicional (Acevedo Álvarez y Tablada Pizarro 2019; Umaña Picado 2022; Vida Estudiantil Universidad Nacional s.f.) y en el rescate de su repertorio asociado (Acevedo Álvarez y Guevara Duarte 2007; Jeremias 2008). Junto a estos esfuerzos, académicos costarricenses han desarrollado investigaciones históricas sobre la marimba tradicional, su música y sus intérpretes; sin embargo, aún faltan estudios sobre el estado actual de esta práctica. En particular, existe un vacío en la investigación sobre el impacto de estas iniciativas de preservación, específicamente en cuanto a sus efectos en la práctica, la enseñanza y el significado cultural de la marimba tradicional.
En respuesta, esta tesis utiliza trabajo de campo para conocer y evaluar críticamente el estado actual de la práctica y enseñanza de la marimba tradicional en dos contextos específicos: comunidades locales (Guanacaste y San José) y programas institucionales de música (escuelas municipales de música y universidades públicas). El estudio examina cómo los métodos pedagógicos, los valores y las prácticas interpretativas difieren y se intersectan en estos contextos. Además, este análisis se sitúa dentro del marco de nociones predominantes sobre la identidad nacional costarricense, un discurso que no solo informa las iniciativas gubernamentales de preservación cultural, sino que también sigue siendo influenciado por el impacto persistente del colonialismo. La tesis explora también cómo aspectos musicales y pedagógicos de la marimba tradicional son moldeados por preocupaciones sociales más amplias, revelando un complejo paisaje cultural en el que su práctica y difusión navegan entre su reconocimiento como símbolo nacional y los desafíos persistentes relacionados con una baja apreciación cultural. Finalmente, esta tesis ofrece una comprensión crítica de las marimbas de concierto, enfatizando los contextos culturales e históricos frecuentemente ignorados en la educación musical occidental. Desafía el distanciamiento de la marimba de concierto respecto de su patrimonio centroamericano, promoviendo una valoración más profunda de las pedagogías orales y las tradiciones culturales que configuraron la identidad de la marimba tradicional.D.M.A
Distributed Optimization Algorithms with Improved Efficiency, Reliability, and Privacy Preservation
Full text linkWith machine learning growing in popularity and relying more on big data and large-scale computations, the demand for distributed computing has continuously increased. This thesis is motivated by the challenges that have emerged as a result of this demand. The first challenge is the spread of computations across multiple processing nodes with varying and unpredictable speeds, leading to per-node delays, known as "straggler" issues. The second challenge is the result of data contribution from sources with diverse privacy requirements, making the development of equitable privacy-preserving approaches for data usage more complex. Motivated by these challenges, the following three research themes form the main contributions of this thesis.
In the first research theme, we incorporate approximation techniques to develop a more efficient coded distributed computing (CDC) approach for mitigating straggler issues. CDC relies on error correction codes to introduce "coded" redundant tasks that are not pure replicas of the original ones. Progress in CDC has mainly focused on the realization of exact computation recovery once a sufficient number of nodes complete their tasks. However, many computational problems involve randomness and are, therefore, naturally tolerant of inexact results. In the thesis, we develop rate-distortion analogs for CDC and design "approximated" CDC schemes in which there are multiple stages of inexact recovery en route to exact recovery. As we will see in the thesis, our schemes help accelerate computation recovery by balancing between accuracy and speed.
In the second and third themes, we incorporate privacy guarantees via differential privacy (DP) to develop new distributed learning methods that are tailored to the varying privacy requirements of participating nodes. Differentially-private federated learning (DP-FL) is an emerging distributed learning framework that enables nodes to collaboratively train a shared model while anonymizing their local data. Progress on DP-FL has mainly relied on implicit assumptions that nodes share the same level of trust with all other participants, have equal privacy constraints, and have fixed incentives to collaborate over time (learning iterations). In the second theme, we extend DP-FL to reflect a setting with multiple, potentially overlapping groups of nodes with varying intra- vs. inter-group levels of trust and analyze how we can control privacy leakage propagation across groups. In the third theme, we show that time variation of the standard DP-FL parameters can improve utility while adhering to personalized privacy constraints. In the thesis, we show our contributions to these two themes, which help incentivize greater participation and enhance participant equity.Ph.D
Climate Variability and Cardiovascular Diseases: The Heart of Oceans
Full text linkBackground: Extreme climates (i.e. cold, heat, rain/snow) can increase the risk of cardiovascular disease (CVD). Climate conditions on land are largely driven by Pacific [El Niño Southern Oscillation (ENSO)] and Atlantic [Atlantic Multi-decanal Oscillation (AMO)] sea surface temperature variability, which have differing regional impacts. Associations between these climate patterns and its extremes on CVD remain understudied.
Objectives: To examine whether: 1) Large-scale climate patterns (ENSO/AMO) are associated with CVD event rates in older adults across North America and UK (Study 1).
2) Regional precipitation is a mediator of such associations (Study 2).
3) Precipitation type (snow/rainfall) is associated with risk of AMI or stroke admissions in the Greater Toronto Area (GTA) (Study 3).
Hypotheses:
ENSO/AMO, warm, and wet conditions will be significantly associated with AMI or stroke event rates among older adults in North America and UK.
Methods & Design: Large population datasets (2000-2019) were used to test each objective, including the Canadian Chronic Disease Surveillance System, US CDC Wide-ranging ONline Data for Epidemiologic Research and National Statistics UK (Studies 1 and 2). Furthermore, administrative databases (Study 3). Primary outcomes were AMI/stroke admissions based on relevant ICD-10/ICD-10CA codes. Monthly or daily region-specific exposure data for sea and land surface temperature/ precipitation were ascertained from satellite data (Studies 1 and 2) and Environment Canada (Study 3).
Overall Results:
Study 1 showed that extreme summer ENSO and AMO were significantly associated with increases in annual AMI events in regions of Canada, the US and UK that experienced warm and wet conditions. Findings from Study 2 suggested that associations between monthly ENSO and AMI in the southwestern US were partially mediated by precipitation. Study 3 found that heavy snowfall is associated with a greater number of daily AMI admissions among older adults in the GTA; while extreme rain is associated with an elevated number of stroke admissions. Temperature was found to be a significant effect modifier.
Conclusions & Relevance:This research demonstrates significant associations between climate variability and CVD outcomes. More extreme conditions are expected worldwide because of ongoing climate change, which may have a substantial impact on CVD risk, particularly in older adults.Ph.D
The sensitivity of fisheries-independent survey indices to decisions of sampling design and intensity and the mitigation of biased precision estimators for systematic sampling
Full text linkFisheries-independent surveys are used to track trends in fish stocks globally. Survey alterations occur for a myriad of reasons (e.g., vessel availability, changes in fish distribution, funding shortfalls). Understanding the sensitivity of survey estimates to changes in sampling is pivotal to sustainable management. We present a case-study of an annual multispecies survey. Simulating distributions for four species with a spatiotemporal model, we evaluated simple random, stratified random and systematic grid sampling designs across four sampling intensities. The systematic design yielded higher-precision estimates at all sampling intensities. However, the commonly used standard error estimator thereof resulted in mean biases from 24% to 63%. We evaluated two alternative standard error estimators that successfully mitigated these biases. Our results indicate that estimates from systematic survey designs may be robust to decreases in sampling intensity, and that analyses such as integrated stock assessments which use these estimates may mitigate model misspecification by applying alternative variance estimators.The presentation of the authors' names and (or) special characters in the title of the pdf file of the accepted manuscript may differ slightly from what is displayed on the item page. The information in the pdf file of the accepted manuscript reflects the original submission by the author
Trans-Threshold Ion Irradiation Damage: Primary and Secondary Damage Mechanisms in Radiation Environments of Emerging Interest
Full text linkMaterials selected for unproven radiation environments require laboratory validation before in-service performance can be confirmed. The most direct validation of performance entails exposure to an identical radiation condition as faced in-service – when this is not possible, extrapolation from benchtop testing to in-service conditions is required. This thesis advances the state-of-the-art in identification of the primary and secondary radiation damage processes required for these extrapolations by application of contemporary experimental and computational methodologies to systems of emerging technological significance. The work is composed of three primary findings chapters. In chapter 4, nanoindentation is used to identify a novel phenomenology of recrystallized tungsten mechanical property evolution in response to hydrogen ion irradiation. Ab initio molecular dynamics illustrate how hydrogen beam energies of 600 eV can result in tungsten primary knock-on atom energies of at least 40 eV. The reduction in nanoindentation yield strength observed at sufficiently high irradiation energies is put forth as a characteristic feature of the production of displacement damage: a new way to measure threshold displacement energy. In chapter 5, the radiation damage tolerance of MoS2 solid state aerospace lubricant is tested tribologically and a damage mitigation strategy proposed based on defect energetics and electronic structure. Chapter 6 introduces the prospect of measuring a low-temperature condensation temperature of helium constrained in metal vacancy clusters, which would help to deconvolute the effect of bubble overpressure driving growth in helium nanobubble systems relevant to fission and fusion technologies. This study of radiation damage provides new insights into mechanisms relevant to nuclear fusion, fission, and deep space environments, and should be of interest to designers selecting materials for these contexts.Ph.D