MSU Libraries Digital Repository (Michigan State University)
Not a member yet
106711 research outputs found
Sort by
A Comprehensive Market Study And Development Of A Novel Pillow Bag Reclosure System
Get PDFThesis (M.S.)--Michigan State University. Packaging - Master of Science, 2025Pillow bag packaging is the dominant format in the snacks category, particularly for chips, due to its cost-efficient production, lightweight structure, and superior barrier properties against moisture, oxygen, and light. Despite these advantages, the lack of effective reclosure mechanisms poses a critical challenge in maintaining product quality upon opening the packaging. This study addresses this limitation by engineering innovative and cost-effective reclosure systems optimized for pillow bag applications. The research encompassed a comprehensive market analysis of existing reclosure technologies, consumer-driven ideation from social media insights, and the adaptation of solutions from various product categories. A diverse range of reclosure concepts were ideated with a focus on usability, manufacturability, and material efficiency. Prototypes of selected designs\u2014Origami Fold, Fold & Tie, and Twist & Tie\u2014were fabricated and subjected to performance evaluation testing. Barrier properties were quantified using the Gravimetric Method under standard and accelerated storage conditions, while Finite Element Analysis (FEA) was employed to investigate shape optimization aimed at reducing stress concentration and preventing progressive tearing in such materials. The selected systems demonstrated significant improvements in barrier performance over conventional open chip packaging, validating their applicability and effectiveness. This research provides a novel, user-centric, and sustainable solution to an enduring challenge in flexible food packaging, advancing both consumer convenience and the industry's sustainability objectives.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Sampling and Selection Bias in Research Using Documented Skeletal Collections
Get PDFThesis (Ph.D.)--Michigan State University. Anthropology - Doctor of Philosophy, 2025Bias research in forensic anthropology has focused primarily on the effects of cognitive bias on laboratory analyses. However, bias not only arises from contextual information at the scene, but can also occur during research endeavors, from design and data collection to analysis and interpretation. To date few studies have focused on potential sources of bias in documented skeletal collections, despite their pivotal role in the production of analytical methods in the United States. Therefore, the purpose of this research is to identify potential sources of sample and selection bias in forensic anthropological research using documented skeletal collections. The three goals of this project address different aspects of sample and selection bias that may be potentially encountered during research development, each addressed in a separate manuscript. The first manuscript uses both demographic and craniometric data from eight United States-based documented skeletal collections to investigate whether specific procurement strategies result in collection specific sample bias. Significant differences between collections were identified in both demographic and craniometric data, indicating the presence of collection-specific sample bias. However, collection-specific sample bias does not obscure patterns of normal human skeletal variation, including sex and population. Therefore, documented skeletal collections remain valid sources of sample data. The second manuscript investigates whether collection samples are representative of the population of interest in forensic anthropology by comparing craniometric data from eight documented skeletal collection to the craniometric data found in Fordisc 3.1.322. Collections were most representative of case populations well-reflected in their samples, such as European Americans who comprise the majority in every documented skeletal collection. Furthermore, inherent sample bias was identified in Fordisc, as nearly one quarter of individuals originate from the same data source; inclusion of those data largely impact model performance. Forensic anthropologists are encouraged to continue case submission to the Forensic Anthropology Data Bank to increase the diversity in Fordisc sample sources, bolster reference sample sizes, and increase the number and variety of the demographic populations represented. The third manuscript uses geostatistical and spatial analyses to investigate whether the physical layout, storage of individuals, and curator involvement at documented human skeletal collections introduce selection bias. Sampling frequency was analyzed using spatial variables, demographic information of individuals within each collection, and interview data, which addressed curator involvement. Selection bias was identified and is mainly attributable to the physical layouts of each collection, including cabinet location and shelf height. However, researcher practices\u2014such as evaluating individuals in chronological order\u2014also contribute to uneven sampling distributions. Researchers need to employ sampling strategies to specifically mitigate the effects of collection layout. Valid methods are required for expert court testimony and all attempts to mitigate bias are essential to upholding ethical standards in the criminal justice system. This dissertation demonstrates sample and selection biases exist within documented skeletal collections, but these collections remain invaluable resources for research in forensic anthropology. Additionally, this dissertation provides a foundation for future investigations into biases for the many documented skeletal collections not included in this dissertationDescription based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Reverse Engineer of Deceptions : Attacks and Defenses for Deep Learning Models
Get PDFThesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025The development of artificial intelligence has been so rapid that every week recently there is a new foundation model update by such as OpenAI, Anthropic, Google, XAI etc.. Ten years ago, most research was still focused on deep neural networks like AlexNet or generative adversarial networks. Now, people talk about large language models (LLM), LLM-based agents, or test time scaling, etc.. However, one thing has remained unchanged: the persistent vulnerability of AI systems to adversarial attacks and backdoor attacks, which threaten their reliability across applications. This dissertation addresses this enduring challenge by advancing the security and robustness of machine learning models through four interconnected contributions. First, it develops a reverse engineering framework to recover original images from adversarial perturbations, enhancing the resilience of image classifiers. Second, it introduces a model parsing technique to infer victim model attributes from attack instances, shedding light on attack transferability and model weaknesses. Third, it examines data poisoning in diffusion models, uncovering bilateral effects\u2014both adversarial vulnerabilities and unexpected defensive benefits\u2014such as improved robustness in classifiers trained on generated data. Finally, it proposes machine unlearning for vision-language models, mitigating harmful outputs and bypassing limitations of traditional safety fine-tuning which relies too much on the spurious correlation. Through all these works, the work tries to reverse engineer the deceptions, delve into the true attributes and methods of the adversaries and then defend accordingly. From image classification to image generation, from classic neural networks to foundation models like diffusion models and vision language models, the work examines through different algorithms and model architectures. These advancements, grounded in rigorous experimentation across diverse datasets, collectively strengthen AI systems against adversarial threats and training-time backdoor injections. The work offers practical tools for secure deployment in high-stakes domains. Beyond immediate applications, this research bridges the gap between the rapid evolution of AI capabilities and the foundational need for trust, laying the groundwork for future investigations into robust artificial intelligence in an era of ever-advancing foundation models.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Racializing Indigenous Society : Native Americans, Euro-Americans, and the Struggle for Authority in Greater Mackinac
Get PDFThesis (Ph.D.)--Michigan State University. History - Doctor of Philosophy, 2025Long before they were incorporated into the fledgling United States, many contested spaces existed throughout North America where Native Americans and Euro-Americans both intermingled and grappled for dominance. A region that this study refers to as Greater Mackinac \u2013 anchored in the Indigenous villages of Mackinac Island, Waganawkezee (L\u2019Arbe Croche), Baawitigong (Sault Ste. Marie), and Mooningwanekaaning (Madeline Island) \u2013 emerged as an arena where Anishinaabeg engaged commercially and diplomatically with Euro-American newcomers while forging kinship ties that gave birth to new peoples. As the United States seized larger swathes of Indigenous Homeland during the nineteenth century, it simultaneously sought to codify Anishinaabeg and their children of mixed ancestry who lived there as racialized "others." This study explores how Greater Mackinac became a meaningful political unit that afforded unique opportunities for Anishinaabe women and their mixed-ancestry kin to broaden the scope of their autonomy and authority. The complex contours of shifting sovereignty in the nineteenth-century Great Lakes Borderlands deserve greater scholastic scrutiny. "Racializing Indigenous Society" contributes to this scholarship by illustrating the extent to which resistance to the American settler-colonial project could germinate in the growing nation\u2019s peripheries. At this study's core are the stories of four Anishinaabe women \u2013 Magdelaine La Framboise, Agatha Biddle, Jane Schoolcraft, and Susan Davenport \u2013 who arose from different nodes of Greater Mackinac to become critical cross-cultural intermediaries. Analyzing their life experiences and those of their interracial families reveals how socially and culturally nebulous spaces could empower Anishinaabe women and their mixed-ancestry kin if they were savvy enough to weave between Indigenous and Euro-American worlds. It also enhances understanding of how mixed-ancestry people came to conceptualize and manipulate their own racial identities during the nineteenth century.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Renewable Energy Landscapes : Mapping the Footprint and Impact of Solar and Wind Energy Development Across the United States
Get PDFThesis (Ph.D.)--Michigan State University. Earth and Environmental Sciences \u2013 Doctor of Philosophy, 2025The transition to renewable energy is essential for mitigating climate change, reducing air pollution, and enhancing energy security. Solar and wind energy with battery storage are now the most inexpensive and broadly available forms of new electricity generation and are thus projected to play a dominant role in a net-zero emissions future. However, solar and wind energy both require the semi-permanent transformation of land, a finite and valuable resource with opportunity costs. The tradeoffs of renewable energy landscapes depend on the siting, construction (and design), and management of the energy development, with key concerns including food security\u2014when transforming agricultural land, habitat loss or fragmentation, reduction in property value, and land degradation. As land-use tradeoffs remain one of the most contentious aspects of renewable energy deployment for supporters and opponents alike, numerous approaches have emerged to mitigate the negative effects of land transformation and enhance the economic and ecosystem services provided. However, the rate of expansion has outpaced the availability of comprehensive spatial and temporal data, leaving knowledge gaps about the footprint of existing landscapes, what practices are being realized and where, and under what policies and incentives can influence the future of renewable energy landscapes. This dissertation addresses these gaps by leveraging remote sensing, geospatial analysis, and publicly available data to quantify the spatial footprint, land use patterns, and tradeoffs of renewable energy infrastructure in the United States (US).Chapter 1 provides a high-level overview of the motivation for this research, and a literature review on ways that ground-mounted solar photovoltaic (PV) installations alter the water cycle at local and regional scales. I compile available knowledge on the approach, scale, climate, limitations, and hydrological outcomes to identify gaps in study design and the data needed to upscale approaches. Together, the literature indicates that solar PV installations unequivocally alter local and regional hydrological processes, and the implications depend on the needs of a region. Importantly, the information needed to replicate many modeling approaches does not exist beyond these few well-studied installations, making it difficult to determine the effects of converting larger portions of watersheds to solar. Chapter 2 addresses this gap, introducing the Ground-Mounted Solar Energy in the United States (GM-SEUS) dataset, the most comprehensive publicly available repository of solar energy infrastructure in the US. Using a combination of high-spatial resolution machine learning and object-based approaches and high-temporal resolution time series segmentation approaches, this dataset cultivates over 15,000 arrays covering 3,000 km2 and nearly 3 million panel-row boundaries covering 470 km2, with standardized and new metadata attributes. Building on existing data sources, GM-SEUS standardizes the definition of an array footprint, improves the spatial accuracy of datasets with uncertainty, and includes under-represented commercial-scale installations, setting the foundation for landscape-scale assessment of the US solar energy landscape. Chapter 3 uses this new dataset to analyze land use and land management conditions of the existing US solar energy landscape through 2024. I find that most solar energy in the US occupies agricultural land and likely lacks a productive vegetation ground cover, presenting opportunities for dual-use practices to have broader adoption. Roughly one quarter of both commercial- and utility-scale solar installations were found to reside on economically marginal land, and commercial-scale installations were preferentially sited on land of significantly low-productivity. Together, these results emphasize the need for stronger policy incentives and improved land management practices to motivate broader development of vegetated and dual-use practices.Shifting to wind energy, Chapter 4 evaluates the agricultural footprint and impact of wind turbines on agricultural production in the US Corn Belt. Wind turbines reduce local crop yields and production, yet the overall effect on food security is small, and decreasing through time. Land leases more than compensate for lost production, increasing incomes for commodity crop farm landowners. These findings demonstrate that wind power can simultaneously support environmental goals and economic security for farmers, without jeopardizing food security. Chapter 5 pulls together the findings of these chapters and reiterates a need for comprehensive and publicly available data on a rapidly expanding renewable energy landscape. With such data, researchers, policymakers, and interest-holders have the opportunity to study the multi-scale tradeoffs and synergies of existing renewable energy land transformation and focus future efforts by investigating what incentives and policies will result in real changes across the landscape.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
IMPACT OF EARLY WEANING, BIOLOGICAL SEX, AND CASTRATION ON ILEAL MUCOSAL TRANSCRIPTOMIC RESPONSES TO LPS CHALLENGE IN PREPUBERTAL PIGS
Get PDFThesis (M.S.)--Michigan State University. Comparative Medicine and Integrative Biology - Master of Science, 2025Adverse experiences that occur early in life, known as early-life adversity (ELA), include factors such as the loss of a parent, exposure to violence, malnutrition, and infections; can disrupt normal developmental processes and significantly increase the risk of diseases later in life. Biological sex is another factor influencing disease risk, with evidence indicating that males and females react differently to ELA. Notably, females tend to be more vulnerable to the adverse long-term effects of ELA. In previous studies conducted by our group, early-weaning stress in pigs\u2014a model for ELA\u2014was linked to a hyperactive enteric nervous system, compromised intestinal barrier function, functional diarrhea, and chronic low-grade intestinal inflammation later in life. These effects were much more pronounced in female pigs compared to castrated male pigs. Still, the exact molecular mechanisms associated with this sexual dimorphism are still not well elucidated. Therefore, the objectives of the present study were to compare how female, male, and castrated male pigs weaned at either 15 days (early weaning condition) or 28 days (late weaning condition) respond to a lipopolysaccharide (LPS) challenge later in life (70-day-old), by evaluating the transcriptome of the ileum mucosa. By comparing samples from female and male pigs, we want to assess the effect of biological sex on the immune response, and by comparing male and castrated male pigs, we want to evaluate how castration status impacts the immune response.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
VOLATILE AND SENSORY PROFILING TO MINIMIZE OFF-FLAVORS IN PULSES
Get PDFThesis (M.S.)--Michigan State University. Food Science - Master of Science, 2025Pulses are a nutrient-dense and sustainable protein source; however, only 17% of Americans consume them at or above the level recommended in dietary guidelines. Incorporating pulse flour into wheat-based products can promote consumption, but adoption remains limited due to undesirable flavors. Volatile organic compounds (VOCs) responsible for these off-flavors primarily include aldehydes, alcohols, ketones, acids, pyrazines, and sulfur compounds. Understanding the factors influencing VOC formation and their impact on sensory perception is critical for improving the quality of pulse-based products. To examine the effects of cultivar, processing (roasting and boiling), and harvest year on the volatile composition of eight pulse cultivars, headspace-solid phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC-MS) was used. Processing trade-offs for mitigating undesirable flavors were assessed by pre-treating some of the pulses by roasting, followed by milling and cooking into model products (porridges), as well as soaking and boiling to evaluate changes in volatile concentrations. Additionally, cultivar differences were analyzed to identify variations in volatile profiles. The impact of crop year was assessed by comparing seven common bean (Phaseolus vulgaris) cultivars grown in Michigan in 2022 and 2023 to one commercially sourced chickpea (Cicer arietinum) grown in 2022. Roasted and non-roasted: flours and model product and boiled pulses prepared from each of the eight cultivars were analyzed by HS-SPME-GC-MS. Hierarchical clustering (HCA) and principal component analysis (PCA) revealed clustering based on harvest year and distinct volatile profiles among cultivars based on seed coat color. Roasting and boiling influenced VOC composition, with variations observed across different compound classes. To further investigate how variations in volatile profiles due to cultivar and processing treatments impact sensory perception, descriptive sensory analysis (DA) was conducted. Given the high cost of sensory panel testing, this study also evaluated the effectiveness of instrumental techniques such as GC-MS and electronic nose (e-nose) in predicting sensory attributes. PCA indicated that sensory variability among cultivars was driven by seed coat color which influenced both appearance and flavor, while HCA indicated that samples with shared sensory attributes clustered based on processing treatment. Pearson\u2019s correlation analysis revealed stronger correlations of e-nose discriminant ions with DA than GC-MS. By integrating sensory and instrumental analyses, this research supports efforts to improve consumer acceptance and increase pulse flour consumption in food products.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Larger sonority difference, larger lag : gestural coordination in speech production
Get PDFThesis (Ph.D.)--Michigan State University. Linguistics - Doctor of Philosophy, 2025Sonority has been one of the most debated concepts in phonetics and phonology. Constraintsinvolving sonority such as the Sonority Sequencing Principle (SSP), the Sonority Dispersion Principle (SDP), or the Syllable Contact Law have long been used by phonologists to understand syllable structure (Sievers, 1881, 1901; Steriade, 1982; Selkirk, 1984; Clements, 1990; Kenstowicz, 1994; Parker, 2002, 2011). However, there is no consensus on the phonetic basis of sonority, either in the articulation or the perception of speech (Albert, 2023). This dissertation explores sonority in speech production.Current speech production theories do not predict the variation of gestural coordination relevantto sonority. However, sonority has been observed to be correlated with systematic variation in gestural coordination, based on CC clusters in Georgian (Crouch, 2022). Also, some observations (Gao, 2008; Shaw and Chen, 2019) suggested that sonority seems to be a factor that systematically correlates to CV gestural coordination variation. In my dissertation, I followed up on these previous studies and explored whether there is a positive correlation between sonority difference and CV lag (the gestural lag between a consonant and a vowel) in English and Mandarin. Based on corpus data of English, as well as Electromagnetic articulography (EMA) experiments participated by English and Mandarin speakers, I found that CV lag positively correlates with CV sonority difference in both languages.There are 32 English stimuli from the Wisconsin X-ray Microbeam Database (Westbury et al.,1990) used to test the main claim in experiment 1. Analyzing the corpus data suggested that there is a significant positive correlation between CV lag and sonority difference. To address the limitation of using an existing corpus and to provide a cross-linguistic comparison, EMA data of 24 English stimuli (experiment 2) and 26 Mandarin tone 4 stimuli (experiment 3) were collected and analyzed. Each set of stimuli in the EMA experiments was read 15 times in different randomized lists. When collecting EMA data, sensors were glued to the tongue tip, tongue blade, tongue dorsum, upper lip, and lower lip of each participant. All the kinematic data were annotated in Matlab using the lp_findgest algorithm of the mview package (Tiede, 2005), where the landmarks were labeled at 20 percent thresholds of peak velocity. The CV lag was computed by subtracting the target onset (onset of gestural plateau) timestamp of the consonant from the target onset timestamp of the vowel (Zhang et al., 2019; Durvasula and Wang, 2023). The sonority difference was quantified by subtracting the C sonority from the V sonority using the sonority scale in Parker (2011). Plots and mixed effects modeling was generated in R (R Core Team, 2017) where CV lag was modeled as a function of the sonority difference, with participant, stimuli, and C duration as random intercepts.The finding is that for all the data, CV lag positively correlates to sonority difference significantly.Sub-groups of the stimuli controlled for consonant place of articulation or vowel height mostly exhibited the expected correlations. I also used consonant displacement and vowel displacement as estimates for jaw movement, and these findings suggest that jaw movement may not be a valid alternative account. The dissertation found a positive correlation between sonority and CV gestural coordination in English and Mandarin. If we make an assumption that larger lags are preferred within a syllable, the finding forms a basis to explain universal constraints such as the SSP and the SDP.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Rationality of Brauer-Severi surface bundles
Get PDFThesis (Ph.D.)--Michigan State University. Mathematics - Doctor of Philosophy, 2025Rationality problems of complex algebraic geometry has a long history. Recent developments, usually referred as specialization method in the literature, have given fruitful new examples of non-rational varieties.We give a sufficient condition for a Brauer-Severi surface bundle over a rational 3-fold to not be stably rational. Additionally, we present an example that satisfies this condition and demonstrate the existence of families of Brauer-Severi surface bundles whose general members are smooth and not stably rational.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references
Moras are unnecessary for stress and other weight-related phenomena
Get PDFThesis (M.A.)--Michigan State University. Linguistics - Master of Arts, 2025The mora is a widely used and highly valuable tool for the analysis of a variety of linguisticphenomena, including but not limited to stress behavior, syllabification, templatic morphology, compensatory lengthening, and tone patterning. Moraic Phonology posits moraic representations in the underlying representations of segments. While the utilization of the mora allows for a breadth of analyses of separate but related processes, the mora also conflates these processes by relating them each to the mora itself. This can be seen as a boon of the theory in that it connects phenomena previously thought unrelated. However, this leads to the challenging task of reconciling a larger set of incorrect predictions of the theory, including moraic mismatches in which the moraicity of segments appears different for separate processes. In the time since Hayes (1986, 1989) it has become clear that many phenomena that are analyzed using moras can be derived computationally \u2013 including via mora assignment/licensing (Crowhurst and Michael 2005; Gordon 2002; Hayes and Wilson 2008; Kiparsky 2003; Zec et al. 2003). This weakens the predictive power of representational mora and assigns more power to computation (of mora assignment or otherwise), without interfering with the status of the mora as the locus of the myriad of different processes for which moras are considered relevant. The use of mora-assigning or mora-licensing processes in computational analyses reduces the strong predictive power of representational accounts, but still conflates many separate processes to be related to the unit itself. For this reason, I suggest a re- examination of the domains which are typically analyzed with moras in order to determine for which processes the mora is a necessary, and preferable tool of analysis. I propose that stress assignment can be computed in a variety of languages without reference to moras or moraic representations. To show this, I build upon Crowhurst and Michael (2005) and Prince (1993) and present an Optimality Theoretic analysis of the stress patterns of several languages which have historically presented issues for moraic analyses, without utilizing moras.Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).Includes bibliographical references