MSU Libraries Digital Repository (Michigan State University)
Not a member yet
106711 research outputs found
Sort by
Knowledge Transfer in Low-rank Matrix Estimation
No full textThesis (Ph.D.)--Michigan State University. Statistics - Doctor of Philosophy, 2025Low-rank modeling plays a pivotal role in signal processing and machine learning, with applications ranging from collaborative filtering, video surveillance, medical imaging, to dimensionality reduction and adaptive filtering.This dissertation explores several questions related to knowledge transfer in low-rank matrix estimation, a central problem in modern statistics with wide ranges of applications. As data from multiple sources become increasingly accessible, new opportunities and challenges arise. We propose methods for leveraging auxiliary information in low-rank matrix estimation, focusing on matrix completion and principal component analysis. All proposed methods are supported by rigorous statistical analysis.We first study the minimax optimal rates in matrix completion. We remove the previously unavoidable logarithmic factor in the convergence rates of several classical estimators, thereby establishing their minimax optimality.Second, we investigate transfer learning for matrix completion. In this problem, data arise from multiple sources and the underlying low-rank structures are similar in nuclear norm. We propose a two-step estimation procedure for the setting where all source datasets are informative. When informative sources are unknown, we develop a data-driven selection procedure.Finally, we study a multi-task learning framework for principal component analysis. We propose a flexible approach for jointly estimating eigenspaces across tasks. Two algorithms are developed under this framework, each achieving a different notion of minimax optimality. The estimators are both adaptive to unknown task similarity structures and robust to arbitrarily distributed outliers.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Effects of Roughness Texture on Turbulent Boundary Layers with Adverse Pressure Gradients
No full textThesis (M.S.)--Michigan State University. Mechanical Engineering - Master of Science, 2025Most wall-bounded turbulent flows, such as those around machined components like ship hulls and propellers, are non-equilibrium in nature due to boundary layers with favorable or adverse streamwise mean pressure gradients. Due to the natural environment in which these machines operate and imperfect manufacturing, these flows often occur over rough surfaces. Many engineering models rely on reduced models of the near-wall layer of flow altered by the roughness (called the roughness sublayer) to provide boundary conditions for the flow above. Studies focusing on characterization of the roughness sublayer for non-equilibrium flow with pressure gradients are limited. This thesis expands on the current knowledge of roughness sublayer characteristics by comparing two flat-plate turbulent boundary layers; one under strong non-equilibrium favorable pressure gradients using existing direct numerical simulation data (Yuan and Piomelli, J. Fluid Mech. 780:192-214, 2015) and the other under non-equilibrium adverse pressure gradients based on a roughness-resolved large-eddy simulation presented herein. The latter utilizes a suction-blowing freestream that induces both adverse and favorable pressure gradients, which induces separation and reattachment of the boundary layer. The focus of this work is on the attached flow regions of both simulations. The roughness sublayer thickness is found to be constant in these regions regardless of pressure gradients. Overall self-similar profiles of mean velocity, dispersive stress, and constant total drag are observed when using a set of sublayer scales ( and , where is the local streamwise mean velocity at the elevation ) for normalization. These statistics suggest that the time mean flow is in quasi-equilibrium regardless of the varying pressure gradients. Additionally, results show that inner or outer scaling within the sublayer fails to accurately capture flow separation or produce self-similar profiles. Leading up to separation, the flow pattern within the roughness sublayer remains self-similar with varying intensities for mean streamwise pressure and velocity despite the presence of strong adverse pressure gradients. Results of this study indicate the potential for extending existing sublayer-unresolved turbulence models to rough-wall non-equilibrium flows and provide wall-friction scaling to assess existing roughness treatments in these models, as far as the roughness Reynolds number is sufficiently high. Future work includes a systematic investigation of the observed self-similarity in flows with different roughness types, pressure gradients, and histories. In particular, the range of conditions where the observed self-similarity breaks down needs to be identified. Funding for this research was provided by the Office of Naval Research.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Probing statistical nuclear properties, the Brink-Axel hypothesis, and astrophysical nucleosynthesis processes with ground and isomeric states in 70Cu
No full textThesis (Ph.D.)--Michigan State University. Chemistry - Doctor of Philosophy, 2025The nucleosyntheis of elements in the universe heavier than iron can be largely attributed to neutron capture nucleosynthesis processes, like the rapid (r), intermediate (i), and slow (s) neutron capture processes. Theoretical models of these astrophysical nucleosynthesis properties require knowledge of astrophysical environments and nuclear physics information like neutron-capture rates and -decay properties. The generalized Brink-Axel (gBA) hypothesis, one of the underlying assumptions involved in theoretical calculations of astrophysical reaction rates, states that the gamma-ray strength function (gammaSF, which is the probability of a nucleus to emit a gamma-ray as a function of energy) is independent of the initial excitation energy and spin of nuclear states in the compound nucleus. Previous investigations into the gBA have confirmed the excitation energy independence of the gammaSF but thus far the spin-independence remains largely unverified. Here, the spin-independence of the gBA was investigated using the beta decay of 70Cu. 70Cu has three beta-decaying states (6- ground state, 3- isomeric state, and 1+ isomeric state), which offer an opportunity to study the gammaSF over different spin ranges at similar excitation energies in 70Zn using the beta-Oslo method. This methodology was also be used to constrain the 69Zn(n,gamma)70Zn reaction rate. In this thesis, results from the first post-trap decay spectroscopy experiments at the National Superconducting Cyclotron Laboratory and the Facility for Rare Isotope Beams are presented. In these experiments, isomer-separated beams of ground and isomeric beta-decaying states of 70Cu were isolated with the LEBIT Penning trap and delivered to the SuN total absorption spectrometer. At FRIB, SuN was upgraded to SuN++. In addition, total absorption spectroscopy results from SuN and SuN++ for each isolated beta-decaying state of 70Cu are presented, as well as their implications on time-dependent energy generation in the astrophysical r process.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
DEFINING THE ROLE OF ENDOGENOUS AND PHARMACOLOGICAL STIMULATION OF NTSR1 ON ENERGY BALANCE
No full textThesis (Ph.D.)--Michigan State University. Neuroscience - Doctor of Philosophy, 2025Obesity results from an imbalance between excessive food intake and insufficient physical activity and energy expenditure. Yet, an incomplete understanding of how the brain modifies feeding and physical activity has hindered the development of pharmacological strategies to support weight loss. The neurotensin (Nts) system is a promising therapeutic target that may be leveraged to promote weight loss in obesity, particularly through the activation of the G-protein-coupled receptor neurotensin receptor-1 (NtsR1). However, it is necessary to determine whether the endogenous systems are sufficient to activate NtsR1 for weight loss and identify which signals from NtsR1 promote anorectic behaviors. The ventral tegmental area (VTA) is well known for containing a large portion of dopamine neurons, most of which express NtsR1 and have been implicated in ingestive behaviors and weight modulation. Subsequently, the central hypothesis of this work is that endogenous and pharmacological stimulation of VTANtsR1 will promote weight-loss behaviors. To further determine whether NtsR1 \u3b2-arrestin signals are responsible for weight loss, we used wild-type mice to deliver systemic or central SBI-553, a \u3b2-arrestin-positive allosteric and G-protein-negative allosteric modulator, and assessed ingestive behavior and metabolic assays. We found that SBI-553 did not affect ingestive behaviors or body weight. By tapping into the lateral hypothalamus (LHA), a brain region rich in Nts, we applied optogenetics in NtsCre to activate the subset of LHANts neurons or their VTA projections. A pharmacological blocker for NtsR1 was administered to determine whether the effects observed were attributable to this Nts-NtsR1 circuit. We found that activating the LHANts-VTANtsR1 circuit can promote locomotion, one of the weight-losing behaviors. Intriguingly, we observed that optogenetic activation of all LHANts neurons or their VTA projections promote physical activity, even in obesity, but had no effects on ingestive behaviors or body weight. Intriguingly, our data also indicates that obesity changes the Nts system via Nts-NtsR1 signal disruption. Altogether, this data supports targeting the NtsR1 in the VTA for weight-loss and will guide the development of therapies to manage obesity.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
QUANTUM-ENABLED HIGH-THROUGHPUT DISCOVERY OF 2D MATERIALS FOR SUSTAINABLE ENERGY : NEXT-GENERATION BATTERIES AND SEMICONDUCTORS
No full textThesis (Ph.D.)--Michigan State University. Materials Science and Engineering - Doctor of Philosophy, 2025Two-dimensional (2D) materials have been Aladdin's cave of condensed matter because they enable the realization of various exotic fields of science. The unexpected behavior of 2D materials has emerged since the first graphene exfoliation from graphite, which grants previously well-studied bulk materials new meanings. Nowadays, 2D materials have contributed to several fields of applications, including electronics and energy storage. However, concern in the development of promising 2D materials originates from experimental approaches, which include the synthesis of qualified materials, the cost of reactants, and the characterization. Here, we combine quantum mechanics, high-throughput methods, and big data to accelerate the exploration of 2D materials, guiding experimental attempts. A huge database of computational 2D materials and their bulk counterparts was first established via density functional theory (DFT). This enables us to study how properties evolve with the bulk-to-2D transition. In addition, we screen the database for a group of materials, which is then demonstrated as an example of thorough data analysis and expansion. In addition to the quantum confinement effect, the highly engineerable structure-property relation of 2D materials allows for the customization of applications. Next, we introduce twistronics, a state-of-the-art engineering of 2D layered materials. The electron density of materials can be modified dramatically after a certain twist between layers is applied, inducing unique characteristics even in homogeneous materials. However, twistronics is highly sensitive to the degree of twist, which demands massive experimental trials. To accelerate the exploration of twistronics, we combine high-throughput codes and DFT to finely construct numerous twisted models and search for unexpected features, respectively. Lastly, we dive into a more functional view of quantum-enabled 2D materials, focusing on how they can tackle the current difficulties in energy applications based on the tunable structure-property relation. For energy storage, we intend to improve the capacity of beyond-lithium ion batteries. Multiple 2D surface modifiers are designed and predicted to stabilize the ion intercalation of lithium alternatives within graphene electrodes. First-principles properties also shed light on the mechanism behind the scenes. For energy conversion, we present an efficient data-driven workflow with DFT, offering unprecedented semiconductors for photo-catalytic water splitting, which can benefit the exploration of 2D materials in the future. This thesis attacks the boundary of 2D materials from the structural and functional perspectives via quantum mechanics. Most significantly, the high-throughput procedures proposed here enable the continuous advancement and development of 2D materials.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Conceptualizing Misconceptions of Rape : Ideology and the Social Context of Sexual Violence
No full textThesis (Ph.D.)--Michigan State University. Philosophy - Doctor of Philosophy, 2025What shapes our understanding of sexual violence such that certain cases, but not others, are recognizable as rape? Why is there confusion about whether a given experience is sexual violence? What do our beliefs about sexual violence imply about our social world? Drawing on socio-political and feminist philosophy, this dissertation argues that the selective recognition of sexual violence stems from an ideological form of social consciousness, for which it develops an account. It further suggests that properly apprehending this form requires situating it within historically specific social relations, which it theorizes through the framework of social reproduction theory.The first chapter develops these guiding questions by analyzing feminist debates on the so-called \u201cgray area\u201d of sexual violence\u2014that is, situations where it is unclear whether an experience constitutes rape. It argues that this ambiguity is a historically specific, social condition affecting our interpretative possibilities, highlighting necessity of feminist critique of conceptions of sexual violence. Responding to this need, Chapter 2 interrogates which approach is best suited to apprehend the historical character of our understanding of rape. Turning to feminist literature on misconceptions of rape\u2014particularly research on rape myths\u2014it argues that existing frameworks are limited either by 1) methodological individualism, or 2) by an inability to account for a range of complex social relations beyond gender and patriarchy, as seen with the radical feminist notion of rape culture as a way of explaining rape myths. It proposes a new taxonomy of rape myths that lays the groundwork for a critical account of their ideological function. Building on this taxonomy, Chapter 3 moves beyond these existing frameworks by offering a marxist ideology critique of rape myths. It identifies their illusory nature, how this relates to their social function, and explains why they are pervasive. Using the concept of hegemony, this analysis revisits the gray area of sexual violations and how rape becomes normalized as part of sex, now understood in ideological terms.Chapter 4 analyzes metaphilosophical debates about the concept of rape and its definitions, challenging the idea that contestation is an essential property of the concept itself. Instead, it shows how contestation reflects specific social and historical conditions, elaborated through anti-racist feminist accounts. Building on the historically specific analysis of sexual violence, the final chapter elaborates on capitalist social relations through the lens of social reproduction theory, while also addressing its limitations\u2014particularly its totalizing account of labor and reductive understanding of sexual violence. By reworking both, it develops a dialectical framework that clarifies how sexual violence and exploitation are mutually constitutive, shaped through the dynamics of social reproduction and subsumption.Together, these chapters underscore the necessity of theorizing sexual violence in relation to its material and historical conditions. Without such analysis, hegemonic perspectives about sexual violence\u2014including those feminist anti-rape activists have sought to displace\u2014risk being reproduced in ways that reify the very social relations from which they emerge.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
YIELDING BEHAVIOR TOWARD PEOPLE WALKING AND BIKING AT ROUNDABOUTS
No full textThesis (M.S.)--Michigan State University. Civil Engineering - Master of Science, 2025Roundabouts have been shown to provide several advantages as compared to alternative forms of intersection control (e.g., signalized, stop-controlled), including reductions in delay, as well as crashes and resultant injuries and fatalities. With that said, an area of concern at roundabouts is the safety of non-motorized users (i.e., pedestrians and bicyclists). An important aspect of this is the yielding behavior of drivers when encountering people walking or biking. This thesis investigates these yielding behaviors with two primary research objectives: 1) to determine which factors influence drivers' yielding decisions at roundabouts; and 2) to determine how best to increase yielding under various scenarios. Data were collected from a series of field observational studies at roundabouts in urban and suburban settings. A series of statistical analyses showed that yielding varied based on factors such as the presence and activation status of RRFBs, crossing direction, roundabout geometry (e.g., number of lanes, entry radius, lane width), and whether the crossing user was a pedestrian or bicyclist.. Furthermore, the analysis showed that yielding rates were significantly lower at multi-lane roundabouts, particularly at exit legs and far lanes, where driver visibility and attention were reduced. Yielding also declined with increased geometric complexity, such as a greater number of approach lanes and wider lane widths. These findings underscore the importance of geometric design interventions\u2014such as tighter radii, reduced lane widths, and improved crosswalk placement\u2014rather than relying solely on signage or public awareness campaigns to improve compliance.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
DEVELOPMENT OF A NEW FATIGUE FAILURE CRITERION USING OPTICAL FLOW-BASED MULTI-GAUGE LENGTH APPROACH
No full textThesis (Ph.D.)--Michigan State University. Civil Engineering - Doctor of Philosophy, 2025Fatigue cracking remains one of the most critical and difficult-to-predict failure modes in asphalt concrete (AC) pavements. Conventional fatigue failure criteria, such as those based on stiffness degradation or stress
7 cycle product, often rely on bulk, fixed-gauge measurements that assume uniform damage accumulation throughout the specimen. However, asphalt concrete is a heterogeneous material where failure typically initiates and propagates in localized regions. This dissertation presents the development of a new fatigue failure criterion based on the onset of localized strain relaxation, captured through a high-resolution, non-contact, optical flow-based strain measurement system.The experimental setup included a customized camera-based measurement system integrated with the Asphalt Mixture Performance Tester (AMPT). The system allowed localized strain tracking by dividing the specimen\u2019s gauge length into multiple segments. Fatigue tests were conducted under cyclic loading, while companion uniaxial monotonic tests were used to study the initiation and propagation of cracks under controlled displacement conditions. Results from monotonic loading revealed that macrocrack formation coincided consistently with the inflection point of the load\u2013time curve, confirmed across three different AC mixtures. Under cyclic loading, a novel failure definition was proposed as the first instance of localized strain relaxation, offering earlier and spatially resolved detection of damage compared to traditional global methods. The new failure criterion was benchmarked against the conventional stress
7 cycle approach. Across all tested specimens, the relaxation-based method identified failure significantly earlier\u2014often two to five times sooner than the traditional approach. This was particularly pronounced in softer mixtures, where the relaxation behavior and strain divergence became more evident. In addition, block-wise strain measurements revealed consistent strain localization patterns, validating the need for spatial resolution in fatigue damage assessment.While the findings demonstrate the robustness of the new failure definition, certain limitations were identified. These include limited camera field of view, and difficulty pinpointing inflection points in soft mixtures with gradual load response. Recommendations include expanding the imaging system to all sides of the specimen, improving synchronization, and testing additional mixtures, including highly modified binders such as HiMA. The developed criterion also presents opportunities for integration into structural models based on the Viscoelastic Continuum Damage (VECD) theory and/or Finite Element Method (FEM).Overall, this study advances the understanding of fatigue failure in AC by redefining it as a spatially localized and temporally trackable event. The proposed framework enhances mechanistic fatigue modeling and supports the development of more resilient, performance-based pavement design strategies.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Gauge Theory and Mass on Riemannian Manifolds
No full textThesis (Ph.D.)--Michigan State University. Mathematics - Doctor of Philosophy, 2025This thesis has two parts, both developing geometric tools for studying the fundamental equations of mathematical physics. In both parts, we focus on the underlying geometric structure, rather than on any particular equations, and we seek to make the constructions broad and flexible.Part I sets up a general procedure for working with the geometric objects in gauge theory\u2013 bundles, connections, and curvature \u2013 that are assumed to be invariant under a symmetry of the underlying space. This is a geometric formulation of what physicists call \u201cdimensional reduction\u201d. It uses symmetry to replace the original gauge fields on an ?-dimensional manifold with a different set of fields on the orbit space. The orbit space has lower dimension, but is singular in general. We focus on collecting results and tools (Section 2.4) that make this approach to variational problems in gauge theory systematic. These results are applied to several specific gauge theoretic PDE\u2019s, including Yang-Mills and Yang-Mills-Higgs equations, and the Yang-Mills flow. Each example consists of a connected smooth manifold together with the action of a compact Lie group, where the action is either transitive or has cohomogeneity-one. In the first case the manifold is a homogeneous space and the original PDE\u2019s equations are reduced to algebraic equations, and in the second case the quotient is one-dimensional and the equations are reduced to a system of ODE\u2019s.Part II studies the analysis over asymptotically flat Riemannian manifolds from the perspective of gauge theory. In Chapters 4 and 5, we define the notion of an asymptotically trivial vector bundle using only the connection; this replaces and generalizes the definition in the literature that uses asymptotic coordinates. We introduce a corresponding gauge group, and a short exact sequence that defines a \u201cfiber over the point at infinity\u201d.Chapter 6 develops theorems about an interesting geometric quantity that emerge from physics: the mass of an asymptotically flat Riemannian manifold. We extend Witten\u2019s approach to the Positive Mass Theorem by developing our own version of recent results on universal positive mass theorems. This is applied to show how the mass is the trace of a bilinear form on the fiber over infinity for any vector bundle associated to the Riemannian structure of the manifold. This approach clarifies the role of the gauge group, and gives a new proof of a result by R. Bartnik that the mass is independent of coordinates. We are also able to obtain a new proof of the positive mass theorem for even dimensional, locally conformally flat or half-flat manifolds.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
The Prophet & The Dream : HistoFuturity as a Metatheory of Educational Change
No full textThesis (Ph.D.)--Michigan State University. Educational Psychology and Educational Technology - Doctor of Philosophy, 2025The purpose of this dissertation is to develop HistoFuturity, an approach to research created by trailblazing author and thinker Octavia E. Butler, as a metatheory for conducting transformative education research. Grounded in my training as an educational psychologist, yet looking beyond it, I propose pathways to just educational futures that transcend disciplinary limits in the face of sociopolitical upheaval. I ask, what would it take for us to enact sustainable change? In response, I examine the entrenchment of research universities in racial capitalist structures, then propose HistoFuturity as an intervention into those structures that allows us to conduct better, more transformative scholarship. I employ archival research, Black feminist speculative praxis, and transdisciplinary reviews of existing literature to formulate a framework that coordinates our education research approaches to critically examine our place in history, while leveraging anticipatory capacity and speculative imagination to envision more beautiful educational futures. In doing so, I argue that a key obstacle to educational change is not merely the absence of equitable, community-engaged research\u2014though more of it is certainly needed\u2014but the lack of visionary coordination across disciplines, methodologies, and communities of practice towards a shared and clearly articulated aim. What I report in this dissertation should be considered the first iterative cycle of metatheoretical development: Drawing on existing scholarship and original research, I propose a framework for structuring education research towards justice and liberation. I then \u201ctest\u201d this framework by discussing it with eight education experts across subfields and career stages\u2014and that\u2019s where this account ends. What comes next is the process of revising and expanding the metatheory based on all I\u2019ve learned and will continue to learn.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references