22689 research outputs found
Sort by
Affine-Diffeomorphic Sequences for Regression and Manifold Learning
Full text linkA multivariate regression model of affine and diffeomorphic transformation sequences is presented, with an adaptation of the model for manifold learning. Leveraging concepts from shape analysis, model states are optimally "reshaped” by diffeomorphisms generated by smooth vector fields during learning. Affine transformations and vector fields are optimized within an optimal control setting, and the model can naturally reduce (or increase) dimensionality and adapt to large data sets via sub-optimal vector fields. An existence proof of solution and necessary conditions for optimality are derived. For the regression model, experimental results on real data sets from the UCI repository are presented, with favorable results in comparison with state-of-the-art in the literature, neural ordinary differential equation models, and densely-connected neural networks in TensorFlow. Experimental results on several toy data sets are also presented for manifold learning, with favorable results in comparison with standard dimensionality reduction techniques
Human Agency in Spinoza: A New Analysis
Full text linkIn this dissertation, I freshly reexamine Spinoza’s account of human agency. On a standardly received scholarly narrative which is most elaborately defended by Olli Koistinen, Spinoza maintains that we are more than bystanders to external forces, even though he rejects free will and accepts causal determinism and necessitarianism. I challenge this orthodox interpretation.
I take the following strategy: firstly, I take a cue from some contemporary attempts to articulate what it is for us to be an agent in which the expression ‘helpless bystanders’ is indeed used as what it is to be contrasted with agents. Specifically, I refer to David Velleman’s account of agency and set it as our point of reference for our default ordinary concept of agency in order to solidify what it is for us to be “more than bystanders to external forces”. Secondly, I show that there are textual reasons to believe that Spinoza cannot make room for what is required of agency on Velleman’s account. Thirdly, I show how Spinoza’s moral theory is designed in such a way that it does not require agency in the sense of the capacity to be more than helpless bystanders to external causes. I then lastly critically engage with attempts to defend the standard reading in the literature according to which we are more than bystanders to external forces, for Spinoza.
The outcome of my doctoral research is that, for Spinoza, we are indeed ‘helpless bystanders’ to the forces external to us insofar as we are ‘acted on’ in his sense, whilst we are still ‘helpless bystanders’ to the forces ‘internal’ to our essence, which in turn is determined by God rather than by our deliberation, insofar as we ‘act’ in his sense. In both cases, we are helpless bystanders to God’s determinations through and through
INNOVATING WHILE BLACK: INVESTIGATING THE BARRIERS AND SUPPORTS THAT IMPACT BLACK SCHOLARS’ SUCCESS
Full text linkThis dissertation presents a multi-paper qualitative study examining how Black scholars navigate the intersecting pressures of race, caregiving, and institutional life. Grounded in Critical Race Feminism (CRF), Role Strain Theory, and Racial Battle Fatigue, the three papers analyze the structural conditions shaping Black academic survival. Paper 1 explores the experiences of Black graduate student parents (BGSPs), highlighting how caregiving responsibilities intersect with institutional neglect to impact mental and physical well-being. Paper 2 examines how Black men in engineering and computing doctoral programs deploy repression as a coping mechanism in response to tokenization and racial isolation, illuminating the psychological toll of survival in predominantly white academic spaces. Paper 3 investigates how Black mothers experience academic institutions as sites of reproductive control and constraint, showing how motherhood becomes a locus of both marginalization and resistance. Together, these studies form the foundation of the Black Resistance and Care in Education (BRACE) model, a recursive theoretical framework that theorizes care labor as strategies of resistance. Methodologically, the dissertation offers a theory-driven secondary analysis of qualitative data informed by critical theories of race. It contributes to scholarship on race in education by illustrating how doctoral education is shaped by systemic exclusion, while also affirming the transformative potential of racially grounded coping and care. Through a layered and intersectional analysis, this dissertation provides a critical roadmap for understanding and supporting Black scholars not just as participants in the academy, but as architects of its reimagination
CHARACTERIZATION AND MORSE INDEX OF MINIMAL SURFACES WITH Y -SINGULARITIES
Full text linkMinimal surfaces, defined by vanishing mean curvature, are fundamental objects in geometric
analysis, with their stability quantified by the Morse index—the number of independent
directions in which the area functional decreases under deformation. While the
theory of smooth minimal surfaces in R3 is well-developed, singular minimal surfaces, such
as those with Y -type singularities observed in soap films, remain less understood. This
thesis advances the study of Y -singular minimal surfaces by investigating their Morse
index, geometric properties, and classification, bridging classical results with modern
geometric techniques.
We introduce a one-parameter family of complete, two-sided Y -singular minimal surfaces
in R3, termed Y -noids, and compute their Morse index as two, with a nullity of
five. As the parameter approaches zero, the family converges to the Y -catenoid union
a plane, we show that the Y -catenoid index is equal to one and the nullity is three.
For rotationally symmetric Y -surfaces with singularities forming a single circle, we calculate
the Morse index by decomposing the problem into a fixed-boundary analysis on
the singularities and a Dirichlet-to-Neumann map for the stability operator, confirming
the Y -catenoid’s index as one. Extending this, we classify complete two-sided Y -surfaces
with a compact interface and Morse index one, proving that a Y -catenoid or a Y -surface
formed with a compact face, an annulus, and an unbounded face have index one. Finally,
we extend Nitsche’s [Nit85] result on the uniqueness of free boundary minimal disk to
the free boundary minimal Y -cones in B3. Specifically, we show that any conformal and
minimal immersion of the standard compact flat Y -cone that meets B3 orthogonally must
itself be a flat Y -cone. Furthermore, we compute the Morse index of the free boundary flat Y -cone, showing that it is two, and we prove that Y -cones are the only free boundary
minimal surfaces of index two in B3.
These results, grounded in minimal surface theory and spectral analysis, enrich the
understanding of singular minimal surfaces’ stability and structure, offering new insights
into their classification and physical relevance
Interpersonal Protective Factors and Mental Health Symptom Self-management among Black Transgender Women: A mixed methods study
No full textBackground: Protective factors are strengths-based characteristics that may decrease the likelihood of experiencing adversity and associated negative outcomes.1 For Black transgender women, adversity can be compounded by intersectional stigma and there is limited exploration of protective factors among Black transgender women. This convergent mixed-methods study was designed to examine protective factors, mental health self-management (MHSM), and symptoms of psychological distress among Black transgender women.
Methods: Survey data was collected nationwide in the United States from 2021-2024. Six multivariate linear regression models were used to test the associations between interpersonal protective factors and current symptoms of psychological distress. Moderation was then tested with the use of interaction terms. Participants from the quantitative sample were purposively sampled to participate in in-depth interviews. Data was analyzed using thematic content analysis and data was then integrated via joint display.
Results: The quantitative sample comprised 151 Black transgender women ages 18-67 (mean=35.9 +/-11). Gender-specific family acceptance (GSFA), positive childhood experiences, and current family social support (FSS) were negatively associated with symptoms of depression and symptoms of PTSD while controlling for covariates. Introduction of interaction terms with MHSM in the models revealed that statistically significant moderation occurred in all associations except PCEs and PTSD. The qualitative sample comprised 28 Black transgender women. Four themes characterizing interpersonal protective factors included Unconditional love, Complicated acceptance, Protective family, and Practical support. MHSM themes included Committing to self, Affirming oneself, Drawing motivation from others, and Leaning on others for support. Integration of MHSM themes with quantitative data revealed contrasts in approaches to MHSM between groups.
Discussion: Participants in this study exhibited interpersonal protective factors, MHSM mechanisms, and effects on psychological distress symptoms in ways that align with what is known in the literature and in aspects that add to the evidence in new ways. This study makes an important contribution to the literature, as it is the first study to examine GSFA and PCEs among a sample of adult Black transgender women. Data in qualitative interviews revealed myriad MHSM strategies that may inform future interventions for Black transgender women and their families to build upon their innate strengths
The Development and Application of Self-Assembled DNA-barcoded ORFeome-scale Human Protein Library
No full textAntibodies play crucial roles in host defense by neutralizing viruses, bacteria, and foreign substances. However, a breakdown in immune tolerance can produce molecules known as “autoantibodies” that mistakenly attack our own tissues, organs, or cells and drive autoimmune pathology. Immunologists have studied numerous autoimmune diseases and identified autoantibodies as biomarkers and pathogenic molecules using traditional biochemical and immunologic techniques, such as western blotting, but these methods are not scalable. Given that the human genome contains approximately 20,000 protein-coding genes, profiling autoantibodies across large patient cohorts remains a substantial challenge.
The advent of nanotechnology, high-throughput biology and DNA sequencing has provided new tools to address this problem. Unbiased screening approaches, such as protein microarrays and phage display, have enabled the identification of novel autoantibody targets. More recently, molecular display techniques such as cDNA displayhave further advanced this effort by enabling the construction of barcoded protein libraries with very high complexity, which can be used to comprehensively profile autoantibody repertoires in patient samples..
This thesis presents the development of Molecular Indexing of Proteins by Self-Assembly (MIPSA) for high-throughput autoantibody profiling. The first part of the thesis describes proof-of-concept studies that characterize DNA–protein conjugates and a barcoded human ORFeome library, and it shows that MIPSA detects autoantibodies against TRIM21 in Sjögren’s syndrome, NT5C1A in inclusion-body myositis, and type I/III interferons in severe COVID-19. Concordance with data from phage immunoprecipitation sequencing (PhIP-Seq) and in-vitro confirmation of the interferon antibodies’ neutralizing activity further validate MIPSA platform.
The second part of this thesis focuses on refining the barcode-human ORF (BC-ORF) pair matching using Oxford Nanopore Technology (ONT) sequencing. We evaluated a set of read aligners including Minimap2, Winnowmap2, NGMLR, and Bowtie2 to assess their ability for ORF recovery from long reads and compared the constructed BC-ORF “dictionaries” to analyze their impact on enrichment analysis from a MIPSA screening data.
Lastly, the third part of this thesis showcases MIPSA’s potential future utility for ligand–receptor studies by examining the binding of epidermal growth factor (EGF) to its receptor, using both EGFR-overexpressing HEK cells and an EGFR-Fc recombinant protein
Understanding EFL Teacher Attrition in K-12 International Schools: Examining Teacher Burnout and Self-Efficacy
Full text linkThis dissertation investigates the factors contributing to high attrition rates among English as a Foreign Language (EFL) teachers in international schools in East Asia, with a focus on teacher burnout and self-efficacy. Framed by Bronfenbrenner’s Ecological Systems Theory and Role Theory, the study explores how cultural, institutional, and psychological dynamics intersect to shape teacher well-being and career sustainability. Utilizing an explanatory-sequential mixed-methods design, the research integrates quantitative data from the Teachers’ Sense of Efficacy Scale (TSES) and Teacher Burnout Scale (TBS) with qualitative insights from semi-structured interviews. Findings indicate that low emotional resilience, unclear role expectations, inadequate institutional support, and cross-cultural tensions contribute to high burnout and intentions to leave, while teachers with higher self-efficacy exhibited more student-centered approaches and adaptive coping mechanisms. Participants emphasized the need for collaborative school cultures, meaningful professional development, competitive compensation, and clearer career pathways. The study highlights the systemic challenges undermining long-term teacher retention and offers recommendations for culturally responsive leadership, supportive peer networks, and well-being-centered policies. These findings contribute to the field of international education by identifying actionable strategies to reduce attrition and support a stable, motivated EFL teaching workforce
Reassessing the CHRIS: An Evaluation of Proposed CA SHPO Online System as an Appropriate Consolidation of the California Historical Resources Information System
Full text linkSince 1972, the California Historical Resources Information System (CHRIS) has served as the official cultural resources inventory for the California State Historic Preservation Officer (SHPO) and the Office of Historic Preservation, but in the 30 year effort by SHPO inventories across the country to digitize their repositories and bring them online to enhance user access and data security, California has fallen far behind much of the nation. After making several attempts to standardize workflows and implement modernization plans since the mid-1990s, efforts are underway to develop a single consolidated digital inventory system called CA SHPO Online.
CA SHPO Online is currently being developed on the Arches heritage database management platform, an open-source software suite developed by Getty Conservation Institute in conjunction with the World Monuments Fund. Arches is built on platform of open-source software tools, including Elasticsearch, PostGreSQL and several OSGeo libraries, a departure from the CHRIS’ present proprietary Microsoft Access and Esri ArcGIS Desktop workflow. With the impending retirement of ArcGIS Desktop by Esri, a migration of the CHRIS’ workflow is necessary to ensure its data security and to avoid disruptions in its service capacity, but is Arches the appropriate platform? Further, is it secure enough to protect sensitive confidential archaeological resource data from unauthorized access while also allowing unrestricted online access to non-confidential historic built environment data?
This study evaluates the history of the CHRIS and its past streamlining efforts, the legal basis for SHPO inventories and their associated access restrictions, the appropriateness of Arches and its open-source components as management systems for confidential data on government servers and what gaps must a digital CHRIS inventory overcome to be successful in the eyes of external stakeholders
NAV1.8+ SENSORY NEURONS PROTECT AGAINST ALLERGIC AIRWAY INFLAMMATION
No full textThe global prevalence of allergic asthma has been steadily increasing, inflicting a heavy burden of disease severity and public health costs. Thus, there is a pressing need for detailed elucidation of its pathophysiologic mechanisms. Allergic asthma is a chronic pulmonary disease characterized by dysregulated type 2 immune responses. While it has long been known that the nervous system initiates cough and bronchoconstriction in asthma, it is increasingly recognized that sensory nerves play an integral role in the immunomodulation of type 2 immunity. Although nociceptors have conventionally been understood to promote allergic exacerbation1, the immune protective aspect of nociceptors is not well understood.
Neuroimmune interactions have emerged as critical regulators of local inflammation. Initially perceived to strictly carry out sensing functions throughout the body, sensory neurons have been found in recent studies to play a role in the maintenance of tissue homeostasis, regulation of metabolism, and the maintenance of barrier immunity. The regulatory ability of sensory neurons highlights a novel paradigm whereby specific nociceptor subsets may protect against type 2 airway inflammation. Studying this neuromodulation addresses an existing research gap and can provide insight into therapeutic approaches to treat allergic asthma.
Here, we show that NaV1.8+ sensory neurons confer a protective phenotype in allergic airway inflammation. By employing the Cre/lox-iDTA system to genetically ablate a particular nociceptor subset in a clinically relevant murine model of allergy, we demonstrate significant suppression of canonical type 2 immunity cellular and molecular mediators. This correlated with a reduction in mucus production, underscoring the extension of protection to physiological features. In probing the underlying mechanism, we found that the protective effect conferred by NaV1.8+ neurons is likely mediated by the neuropeptide CGRP, a known negative regulator of ILC2s. Moreover, the effect of CGRP on ILC2s is likely a direct one, indicated by the lack of involvement of alarmins, which are strong activators of ILC2s. Collectively, these results demonstrate that NaV1.8+ sensory neurons induce a protective phenotype against both innate and adaptive players of type 2 airway inflammation through the suppression of eosinophils, monocytes, Th2 cells and ILC2s by direct engagement of the neuropeptide CGRP
Asymptotics and Scattering for Relativistic Field Equations
No full textRelativistic field equations play a fundamental role in both mathematics and physics, as they govern the dynamics of spacetime and fields. This thesis focuses on the long-time dynamics and scattering theory of several relativistic field equations in 1+3 dimensions.
The Einstein equation lies at the core of the general relativity. One of the most significant and challenging problems is the stability of known stationary solutions to the Einstein equation. Among them, the most fundamental case is the stability of the Minkowski spacetime. This highly nontrivial problem is first solved in the monumental work of Christodoulou—Klainerman. An alternative and much simpler proof was given by Lindblad-Rodnianski, whose approach has also been generalized to many Einstein-coupled systems by later works. In this thesis, we present a proof of the stability of Minkowski for the Einstein equation coupled with the massless Dirac equation, which has not been studied before.
We show that this coupling, while causing more subtle behaviors than the Einstein vacuum or Einstein—(massless) scalar field system, still admits a global stability result.
For the coupling with a massive field, the analysis changes drastically due to different dispersion relations and the breakdown of scaling symmetry. One of the most fundamental problems in this direction is the Einstein—Klein—Gordon system, i.e. the coupling with a massive scalar field. Several previous works developed different approaches to deal with this problem and proved the stability of Minkowski for the Einstein—Klein—Gordon system. In this thesis, we further study the physical-space scattering theory of systems of this type.
We show two types of results: The asymptotic completeness (forward expansion of the solution), and scattering from infinity (existence of global solutions with prescribed asymptotic expansion). We first prove these results for a wave—Klein—Gordon system that models the Einstein—Klein—Gordon equation, and then for the massive Maxwell—Klein—Gordon equations. Previously, these physical-space results have only been obtained for massless coupling systems