MSU Libraries Digital Repository (Michigan State University)
Not a member yet
106711 research outputs found
Sort by
CREATING ACCESSIBLE UML CLASS DIAGRAMS
Get PDFThesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025Unified Modeling Language (UML) Class Diagramming is the commonly accepted mechanism used to describe relationships between software components. In addition, it is an essential educational tool that is used to convey the structure of software and the patterns of software design to students. Unfortunately, UML is a visual-only mechanism and therefore is not useful for developers and students who are blind or have visual impairments. This work describes a method for conveying class diagrams using audio, which addresses this lack of a tool to support these populations. This method works by dividing the views of a diagram into smaller spaces. Elements in these subspaces are conveyed through manipulation of audio properties. Multiple user studies were performed to prove that the tool is viable for conveying the static structure of software elements and that the workload required to use the tool is reasonable. The results of the studies indicate that the tool is effective and requires only slightly higher mental workload than traditional class diagrams.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Proactive Schemes : Adversarial Attacks for Social Good
No full textThesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025Adversarial attacks in computer vision typically exploit vulnerabilities in deep learning models, generating deceptive inputs that can lead AI systems to incorrect decisions. However, proactive schemes approaches designed to embed purposeful signals into visual data can serve as \u201cadversarial attacks for social good,\u201d harnessing similar principles to enhance the robustness, security, and interpretability of AI systems. This research explores application of proactive schemes in computer vision, diverging from conventional passive methods by embedding auxiliary signals known as "templates" into input data, fundamentally improving model performance, attribution capabilities, and detection accuracy across diverse tasks. This includes novel techniques for image manipulation detection and localization, which introduce learned templates to accurately identify and pinpoint alterations made by multiple, previously unseen Generative Models (GMs). The Manipulation Localization Proactive scheme (MaLP), for example, not only detects but also localizes specific pixel changes caused by manipulations, showing resilient performance across a broad range of GMs. Extending this approach, the Proactive Object Detection (PrObeD) scheme utilizes encoder-decoder architectures to embed task-specific templates within images, enhancing the efficacy of object detectors, even under challenging conditions like camouflaged environments.This research further expands proactive schemes into generative models and video analysis, enabling attribution and action detection solutions. ProMark, for instance, introduces a novel attribution framework by embedding imperceptible watermarks within training data, allowing generated images to be traced back to specific training concepts\u2014such as objects, motifs, or styles\u2014while preserving image quality. Building on ProMark, CustomMark offers selective and efficient concept attribution, allowing artists to opt into watermarking specific styles and easily add new styles over time, without the need to retrain the entire model. Inspired by the proactive structure of PrObeD for 2D object detection, PiVoT introduces a video-based proactive wrapper that enhance action recognition and spatio-temporal action detection. By integrating action-specific templates through a template-enhanced Low-Rank Adaptation (LoRA) framework, PiVoT seamlessly augments various action detectors, preserving computational efficiency while significantly boosting detection performance. Lastly, the thesis presents a model parsing framework that estimates "fingerprints\u201d for the generative models, extracting unique characteristics from generated images to predict the architecture and loss functions of underlying networks\u2014a particularly valuable tool for deepfake detection and model attribution. Collectively, these proactive schemes offer significant advancements over passive methods, establishing robust, accurate, and generalizable solutions for diverse computer vision challenges. By addressing key issues related to the different vision applications caused by conventional passive approaches, this research lays the groundwork for a future where proactive frameworks can improve AI-driven applications.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Holding on to kinship in a global neoliberal world; personhood, place, and affect in everyday life in Lahore, Pakistan
Get PDFThesis (Ph.D.)--Michigan State University. Anthropology - Doctor of Philosophy, 2025Scholars agree that improvements in communication and transportation technologies, and the integration of national economies into a global market in new ways, characterize the distinctive era of globalization. The changing scale, volume, and velocity of global connections have transformed even the minutiae of everyday local life, so that the global and local are embedded in the ways people live out their daily lives and make decisions pertaining to it. In trying to map out how the global mobility of people in urban Pakistan, in both physical and virtual ways, affects their sense of place and personhood, this dissertation conducts an ethnographic investigation of the transnational social relations and narratives of people living in Lahore. Speaking with varying demographics in Lahore, it explores the aspirations and motivations of people in urban Pakistan that shape their desire to go abroad or stay in Pakistan, and the effects on people in Pakistan of the migration of relatives abroad. By delineating these aspirations, motivations, and effects, this dissertation brings to light the friction between global and local ways of being and how that tension is experienced differentially in line with factors such as gender, class, and generation. It studies transnational social relations, material exchange, and digital communication among people in Lahore and their relatives abroad to understand how these practices enable them to maintain a sense of place and personhood, even as they simultaneously also shape a politics of place and belonging, in an imperial global neoliberal order. This dissertation takes inspiration from anthropological, sociological, philosophical, and cultural studies theoretical frameworks of place, personhood, and affect, and builds upon scholarship in transnational migration, globalization, and kinship studies. It illustrates the operation of kinship and affect as people in Pakistan, and some of their diasporic counterparts, try to hold onto a sense of place and personhood in a globalizing world.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
TRANSACTION COSTS IN CONSTRUCTION PROJECT TEAM COMMUNICATIONS : LANGUAGE MODEL AND NETWORK SCIENCE APPLICATIONS
No full textThesis (Ph.D.)--Michigan State University. Planning, Design and Construction - Doctor of Philosophy, 2025Project teams with diversified interests and talents allocate their limited resources according to a shared form of a plan to constitute temporary alliances in Architectural, Engineering, and Construction (AEC) industry. These resources, or goods, service, and knowledge, are transferred across technologically separable interfaces and those activities for which firms provide less costly management can be organized within productive teams. The dynamic nature of acquiring and processing information, contractual and organizational relationships, technology adoption and uncertain environments affect the collaboration dynamics which lead to determining overall performance of the projects. However, despite the critical role of these \u2018softer\u2019 social dynamics, they have yet to be comprehensively examined through the lens of transaction cost (TC) theory in existing literature, especially amid profound technological and environmental shifts.This dissertation examines these dynamics and the potential of generative AI (GAI), specifically attention-based language models, within inter-organizational project networks to uncover tangible lessons learned for improving project collaboration. It outlines the steps from data acquisition to analysis, focusing on multi-level communication dynamics and performance-related documents from a complex healthcare project and a medium-complexity mixed-use project in Michigan, USA. The study integrates quantitative and qualitative datasets, including emails, semi-structured interviews, surveys, construction documents, and interorganizational meeting recordings. Social Network Analysis (SNA) guides the investigation of communication networks, while language models are assessed for classification and communication-related tasks. Programming languages Python and R facilitate data cleaning, statistical tests, and visualizations, with code snippets provided in the appendix for replication. After the introductory section, the second chapter (1) investigates how complex relationships from task descriptions can be categorized and tracked, comparing current machine learning methods and fine-tuned open-source language models with limited training data, considering real-world AEC industry mindset. (2) The third chapter explores how times of disruption (ToD) affect communication dynamics among interorganizational project networks through a COVID-19 case study. (3) The fourth chapter explores text complexity metrics, absorptive capacity (ACAP) measured through tier and role based characteristics, and GAI\u2019s role in AEC communications, analyzing how project parties perceive its effectiveness in refining emails and RFIs to reduce TCs. Key findings highlight the value of GAI and strategic communication: (1) Fine-tuned language models can outperform current machine learning methods in categorizing complex tasks, especially when provided with detailed descriptions that reveal underlying social dynamics. (2) Balancing direct and indirect communication can enhance information flow between central project members and distant parties, particularly subcontractors with high specificity, during ToD. TC Theory highlights the need for adaptive, tier-specific management, and flexibility, especially from Tier 1 leaders, to sustain network stability and project continuity. (3) GAI\u2019s editing capabilities can effectively tailor text complexity and readability to facilitate project communication, provided appropriability safeguards private information. (4) Overall, a cohesive strategy of strategic communication, adaptability, and generative technologies can optimize project outcomes by reducing knowledge transfer frictions and harnessing stakeholders\u2019 authentic strengths.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Pathogen Classification Using AI-Enabled Hyperspectral Microscopy to Detect Biological Variations
Get PDFThesis (M.S.)--Michigan State University. Biosystems Engineering - Master of Science, 2025Timely detection and identification of pathogens are crucial for safeguarding food safety, public health, and environmental monitoring. However, conventional methods often struggle to account for subtle biological variations in microbial physiology and the nuanced differences among closely related serovars. The objectives of this work were to: i) systematically review AI applications for imaging-based pathogen detection under stress conditions; ii) develop an AI-enabled hyperspectral microscopy framework for rapid detection of stressed cells under low-level antimicrobials; iii) enhance data processing of this framework to classify Salmonella serovars. Critical gaps identified from the systematic review highlighted limited research on stressed pathogen detection and inconsistent reporting of laboratory protocols and data pipelines. An optimized laboratory protocol captured distinct hyperspectral profiles of E. coli K-12 in both normal (i.e., viable and culturable) and viable but non-culturable (VBNC) states induced by low-level antimicrobials (n = 200). An EfficientNetV2 convolutional neural network (CNN), trained on pseudo-RGB images from these hyperspectral data, achieved a 97.1% accuracy for VBNC classification, outperforming standard color images. Additionally, hyperspectral data detected subtle biological differences of five Salmonella serovars, i.e., Kentucky, Johannesburg, Infantis, Enteritidis, and 4,[5],12:i:- (n = 500). Enhanced data preprocessing techniques and multimodal fusion methods were introduced, incorporating both spectral features (via manual feature selection vs. data-driven feature extraction) and spatial features (CNN-based). Data-driven feature extraction outperformed manual selection, and multimodal fusion further improved classification accuracy to 82.40%. These findings demonstrate that integrating hyperspectral microscopy with AI-enabled data analysis enhances classification capabilities, paving the way for practical, streamlined rapid pathogen detection solutions.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
STREAM FISH RESPONSES TO CHANGING ENVIRONMENTAL CONDITIONS : USING A FUNCTIONAL BIOGEOGRAPHY APPROACH ACROSS BROAD SPATIAL EXTENTS
Get PDFThesis (Ph.D.)--Michigan State University. Fisheries and Wildlife - Doctor of Philosophy, 2025Stream habitats and the fishes they support are threatened by environmental factors operating within catchments, and efforts to contextualize species responses to those factors are essential to design and implement effective management actions. Investigating the morphological, physiological, phenological, or behavioral characteristics of species can improve mechanistic understanding of how stream fishes respond to environmental factors. Insights gained from such traits-based investigations can be used to predict changes in the structure and function of stream fish assemblages across space and through time. Despite the utility of traits-based investigations, understanding how environmental factors influence functional traits of stream fishes across broad spatial extents (e.g., continental) remains incomplete. Few traits-based investigations have been conducted at such broad extents because availability of standardized datasets representing distributions of stream fishes is often limited, as are datasets characterizing environmental factors consistently over large regions. Therefore, the goal of my dissertation is to use a functional biogeography approach to describe, explain, and predict functional responses of stream fishes to changing environmental conditions at a continental extent. In my first chapter, I use RLQ and fourth-corner analyses to describe relationships between 17 environmental variables and 16 traits for 597 stream fish species within the conterminous United States. I evaluate the generalizability of trait-environment relationships across the study region and show that while the strength and multivariate structure of trait-environment relationships vary, some relationships, including positive associations between migratory species and forested land cover, were significant in multiple ecoregions. In my second chapter, I investigate consequences of biodiversity change on the stability of stream fish metacommunities throughout the conterminous United States. To do so, I develop a structural equation model to integrate predictions of alpha, beta, and gamma diversity, functional redundancy, and compositional and functional variability at local to regional spatial scales. I show that multiple forms of biodiversity contribute to the compositional and functional variability of stream fish metacommunities and generate insights into how local and regional management efforts may help to achieve sustainable outcomes for freshwater ecosystems. In my third chapter, I develop a decision-support framework to promote the use of ecological thresholds (which represent the intensity of a human landscape stressor that leads to a severe decline in fishes) in decision-making applications. This framework integrates threshold status indices, summaries of protected area distributions, and knowledge of multiple stressor configurations within stream catchments to inform conservation and restoration efforts for nearly 1.73 million catchments located throughout the conterminous United States and Europe. The findings highlight the pervasive influences of agricultural land use on stream habitat and indicate that widespread degradation may result from increased urban development within catchments that are poorly protected. Collectively, my dissertation contributes to an improved understanding of how the functional characteristics of stream fishes vary across space and through time and demonstrates the value of using traits-based approaches to characterize vulnerability of stream fishes to human landscape stressors at broad spatial extents. The insights generated in my dissertation can be applied to support management decision-making processes to help limit further degradation of stream habitat and contribute to addressing the global freshwater biodiversity crisis.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Graph-based learning for Community Detection and Hub Node Identification
No full textThesis (Ph.D.)--Michigan State University. Electrical and Computer Engineering - Doctor of Philosophy, 2025Many real-world systems can be represented using complex networks, where the different agents and their relations are represented as nodes and links, respectively. Traditional network models employ simple graphs where the graph is represented by a set of vertices and edges that connect them. With the advances in data acquisition technologies and the different types of data that are available, the simple graph model becomes insufficient to describe the higher dimensional relational data sets. For instance, in social networks, users can be defined as nodes with multiple types of interactions like friendship, collaboration, and economic exchange, connecting them. Furthermore, each node is often associated with attributes such as demographics or interests. To overcome the limitation of existing simple graph models, multi-dimensional graphs such as multiplex networks have been proposed. Similarly, in order to capture the node information that is available in most real-world networks, attributed graphs have been introduced.Given a large scale complex network, one is usually interested in learning the underlying graph structure, such as the community structure or hub nodes. These structures uncover meaningful patterns and provide insights within complex networks. Community detection identifies groups of nodes that are more densely connected to each other than they are to the rest of the network. Hub nodes, on the other hand, correspond to nodes which are densely connected to the rest of the graph and play a critical role in information processing in the network. Although there are numerous works on community detection in single-layer networks, existing work on multiplex community detection mostly focuses on learning a common community structure across layers without taking the heterogeneity of the different layers into account. Beyond detecting communities within a single multiplex network, many applications may require comparing the community structures of two or more multiplex networks. Furthermore, most of the existing community detection methods focus solely on the graph connectivity information. In attributed graphs, where each node is associated with an attribute vector, the community detection methods that focus only on the edges and the data clustering methods that focus only on the attributes of the nodes become insufficient. Traditional hub detection methods rely mostly on graph connectivity without taking the node attributes into account. This thesis addresses the limitations of learning these graph structures in high-dimensionaland attributed networks. Novel algorithms for multiplex and attributed community detection, as well as approaches for discovering discriminative communities between two multiplex networks, are introduced using graph spectral theory and graph signal processing methods. Similarly, a graph signal processing approach that takes into account both the graph topology and node attributes is introduced for hub node identification.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
IDENTIFYING DISCREPANCIES BETWEEN INWARD AND OUTWARD ELECTRON TRANSFER IN SHEWANELLA ONEIDENSIS
Get PDFThesis (Ph.D.)--Michigan State University. Biochemistry and Molecular Biology - Doctor of Philosophy, 2025Addressing the climate crisis requires not a single breakthrough, but a suite of well-understood, adaptable solutions spanning technology, policy, and practice. To limit warming to no more than 1.5\ub0C above pre-industrial levels, we must approach net zero carbon dioxide (CO2) emissions around the mid-twenty first century. However, achieving this will require a coordinated effort across multiple disciplines, with combinations of biology and technology playing a crucial role in supporting these initiatives. One promising biotechnology, microbial electrosynthesis (MES), has the potential to significantly reduce net CO2 emissions if implemented on an industrial scale. In future MES systems, microbial species capable of extracellular electron transfer (ET) and carbon fixing reactions could recycle CO2 from industrial emissions directly into useful organic molecules. While MES and its key components (the bacteria-electrode interface and ET), could become valuable tools in the broader effort to lower net CO2 emissions, fundamental questions remain for even the most well-understood extracellular ET pathway, the Mtr pathway. The Mtr pathway is the metal reducing pathway from Shewanella oneidensis, a bacterium that can use extracellular electron acceptors when the available oxygen is insufficient for respiration (outward ET). The Mtr pathway is also bidirectional, an important feature for a model organism used to study MES. Because MES requires a robust bacteria-electrode interface for electron transfer into the bacterium (inward ET), the bidirectionality of the Mtr pathway provides an excellent vehicle for studying the mechanisms and bottlenecks that constrain inward ET in S. oneidensis or comparable systems. Despite the established bidirectionality of the Mtr pathway, there is a persistent asymmetry between outward and inward ET, with outward electron transfer being consistently higher in magnitude. Tefft and TerAvest (2019) developed an S. oneidensis strain expressing butanediol dehydrogenase (Bdh), a non-native NADH-dependent enzyme. The enzymatic reaction Bdh catalyzes, acetoin reduction to 2,3-butanediol, can act as an indicator of electron transfer to cytoplasmic carriers via NADH dehydrogenases. However, this direction is the opposite of the respiratory direction and is thermodynamically limited for inward ET. In Chapter 2, I use a thermodynamic model to compare inward ET through S. oneidensis NADH dehydrogenases under three energetic coupling scenarios. I also use a complementary experimental approach to assess qualitative changes in membrane potential at the single cell level for electrode-attached S. oneidensis. In Chapter 3, I compare the extracellular component of inward and outward ET by using two thermodynamically favorable ET paths. Under conditions with and without supplemental flavin (a known ET mediator for S. oneidensis), I use both chronoamperometry and cyclic voltammetry to conclude that inward and outward ET occur through different mechanisms for anaerobic S. oneidensis. In Chapter 4, I follow up on an unpublished observation that growth medium impacts inward ET performance in bioelectrochemical systems (BESs) with minimal medium lacking a carbon source. Specifically, when S. oneidensis was pre-cultured in minimal medium rather than rich medium, inward ET ability increased. I performed differential protein analysis and found that pre-culture in minimal medium appears to prime S. oneidensis for inward ET more effectively than pre-culture in rich medium. Growth in minimal medium made proteins in energy conserving pathways more abundant, and proteins involved in translational processes less abundant. Together, Chapters 2, 3, and 4 describe bottlenecks along the inward ET pathway that, if alleviated, could lessen or eliminate the discrepancy between inward and outward ET rates in S. oneidensis.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
DATA, MACHINE LEARNING, AND POLICY INFORMED AGENT-BASED MODELING
No full textThesis (Ph.D.)--Michigan State University. Computational Mathematics, Science and Engineering - Doctor of Philosophy, 2024Agent-based models (ABMs) examine emergent phenomena that arise from individual agent rules. This work extends the basic ABM paradigm in three key areas: data integration, evaluation of policies, and incorporating machine learning techniques. The dissertation investigates how data-driven approaches can enhance the accuracy of ABMs, explores the practical applications of ABMs in developing policies for real-world issues, and examines the fusion of machine learning with ABMs to optimize model design and functionality.The dissertation begins by establishing the background and fundamental principles of agent-based models, highlighting their evolution from simple cellular automata to intricate systems encapsulating decision-making and adaptive behavior. It examines the role of these models in simulating dynamic interactions within systems, especially in scenarios where traditional methods may fall short of capturing the complexities of agent interactions. Following the introduction of key concepts, a series of projects that function in pairs demonstrate the versatility of ABMs and address the three key areas discussed above.The first pair addresses data integration of GPS deer movement data into a generalized Langevin model and its use in uncertainty quantification of disease spread. Exploratory data analysis revealed a discernible non-parametric trend in the GPS data with non-Gaussian statistics. This analysis led to a model that is consistent with the observed data. Subsequent incorporation of chronic wasting disease (CWD) and population dynamics were used to forecast the prevalence of CWD. This extended model was analyzed with a global sensitivity analysis that tied variance in disease prevalence to variance in the parameters of the model, providing predictions of future prevalence of the disease.The second pair examines policy evaluation, specifically strategies for mitigating disinformation in social networks. Multiple strategies were evaluated on various topologically diverse networks that led to policy recommendations. Simulations on these graphs revealed challenges associated with large network simulations, particularly in computational cost and influence of network topologies. These challenges led to a method to miniaturize real social networks while preserving key attributes, enabling more efficient and realistic simulations to run on artificial social networks.The final pair investigates the possibility of inverting the ABM paradigm to instead have agents learn their own rules through environmental interactions. Reinforcement learning was applied to a model of conflict based on capture the flag, where an agent learned in progressively difficult competitions. The emergence of deterrence was explored through adding asymmetries between competing teams, and differential equation-based models were created to help interpret results.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
SYNTHESIS AND ENGINEERING STUDIES OF BIOBASED AND BIODEGRADABLE- COMPOSTABLE POLYMERS
No full textThesis (M.S.)--Michigan State University. Chemical Engineering - Master of Science, 2024In this thesis, the focus is on Polylactide (PLA) based polymers, recognized as the leading 100% biobased resin globally. These polymers align with the Circular Economy model proposed by the Ellen MacArthur Foundation, offering both composting and recycling as viable end-of-life options. PLA is commercially produced by transforming lactic acid into lactide, followed by polymerization. The unique stereochemistry of PLA molecules significantly influences manufacturing processes, performance characteristics, and overall processability. Despite its importance, the role of stereochemistry in shaping PLA's product performance and applications is often overlooked and not thoroughly understood. The present work discusses stereo-complex PLA using reactive extrusion to improve the mechanical performance of thermoplastic PLA matrix.It also focuses on polymerization of meso-PLA and further modification of Poly (meso-lactide) and commercial PLA grades for applications in paper- coating. Lastly, reactive extrusion is emphasized as a continuous and facile method for synthesis of PCL and PLA. The overall aim of the thesis is to focus on utilization of PLA in different domains of application through synthesis or modification using twin-screw extruder as a facile approach.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references