20566 research outputs found
Sort by
An Instructional Remodel: How Adopting a Unifying Model-of-Instruction Impacted Teacher Self-Efficacy and Teacher Development and Evaluation Efforts in a Midwestern High School
Get PDFThis mixed methods study sought to determine the effect of adopting a unifying model-of-instruction on teacher self-efficacy and the development and evaluation processes of administrators. By assessing the level of teacher self-efficacy (Tschannen-Moran & Hoy, 2001) in the nascent stages of the instructional model’s implementation, then reassessing it after a multiple-month sequence of professional learning, the researcher provides a judgment on statistical significance concerning the two measurements and conjectures the extent to which the instructional model impacted any differences in self-efficacy levels. Using two separate “before and after” interviews with purposefully selected administrators, the researcher explored how the unification of instructional practice with a common framework for providing feedback impacted school leaders as they sought to develop teachers around a consistently defined paradigm in this model-of-instruction
Adversarial Machine Learning: Methods for Attacks and Defenses
Get PDFWith the rapid development of machine learning in real-world applications, enhancing security plays an important role. Adversarial machine learning focuses on understanding malicious actions from attackers and developing defensive techniques against such threats when deploying machine learning systems. An attack can occur in different scenarios, such as poisoning attacks during the training stage and evasion attacks during the testing stage. Although extensive research has explored defense strategies to deal with these harmful attacks, there is a need for further research into areas such as how to counteract malicious attacks with healthy noise or how to train an adaptive defense against a mixture of attacks. Additionally, developing novel attack methodologies is essential for uncovering underexplored vulnerabilities in the training and testing pipelines, thereby providing defenders with deeper insights into the inherent weaknesses of model architectures. Most adversarial attacks primarily aim to degrade overall classification accuracy; however, there is a notable lack of attack strategies that target models designed for fair prediction or multimodal retrieval. Furthermore, malicious users continuously devise subtle methods to disseminate harmful content on social media, necessitating the development of intelligent systems capable of detecting and mitigating such content. Vision-Language Models, which have been widely used in real-world applications, hold significant potential for fostering safer and more respectful online environments.
The goal of this dissertation is to address critical challenges in ensuring safety in machine learning models, focusing on the development of novel attack and defense methods. We begin by investigating two defenses against specific attack types: poisoning attacks and evasion attacks. Next, we examine the potential threats posed by adversaries targeting the fairness of machine learning models, introducing a novel poisoning attack on fair machine learning systems. We then analyze the vulnerabilities of multimodal pre-trained models under adversarial attacks. Finally, we explore methods for detecting and mitigating hateful content in multimodal memes utilizing Vision-Language Models. In this dissertation, we present the following frameworks and algorithms. We develop a defense against data poisoning attacks by leveraging the influence function, which helps the model reduce the harmful effect of poisoned training data; We introduce a defense against evasion attacks via adaptive training, which makes the model adaptive and robust to unseen attacks at the testing stage; We design an attack on fair machine learning models, which not only degrades model accuracy but also hinders the model\u27s fairness objective; We investigate adversarial attacks that degrade the multimodal retrieval capability of pre-trained models; We introduce a definition-guided prompting-based method for detecting hateful memes; We develop a unified framework to transform hateful memes into non-hateful versions
An Examination of the Relationship Between Postsecondary Education and Earnings in the Architectural Woodworking Industry
Get PDFThe architectural woodworking industry offers a potential option to job seekers who are looking for alternatives to a formal 4-year postsecondary education. This preregistered quantitative study explores the relationship between postsecondary education and earnings for employees within the architectural woodworking industry. Framed by Human Capital Theory, the research investigates three questions: (1) do postsecondary credentials account for additional variance in income within the industry, (2) is the relationship between postsecondary education and income moderated by whether the credential is related to woodworking, and (3) do individuals with specific types of postsecondary credentials earn higher incomes than those without postsecondary credentials? Data is collected from full-time employees within architectural woodworking companies. The study aims to add to the vast body of empirical evidence on the postsecondary education and earnings relationship, contributing to a gap in the literature on the relationship within the architectural woodworking industry and potentially other specialty trades-based industries. The study will also provide evidence to inform individuals considering the architectural woodworking industry as a career path
An Empirical Assessment of Cannabidiol, Stress, Anxiety, and Sex Differences: A Randomized, Placebo-Controlled Trial
No full textCannabidiol (CBD), a non-intoxicating molecule derived from the Cannabis sativa L. plant, displays broad therapeutic potential in the context of anxiety-related vulnerability. Very little work has directly examined the effects of CBD on stress among human subjects; however, existing data suggest CBD has the potential to reduce stress. Two studies have examined the acute versus repeated effects of CBD and reported null effects. The goal of the current study is to replicate and extend existing evidence suggesting CBD reduces stress by examining the effects of acute administration of 300mg CBD compared to repeated CBD administration at the same dosing level (verses placebo) on physical symptoms of stress, perceived stress, physical symptoms of anxiety and state anxiety. Theoretical and empirical evidence suggest the potential for sex differences with respect to CBD administration. The current study also sought to explore potential sex differences in this context. Participants were 79 individuals (Mage = 22.70; SDage = 7.56) self-reporting elevated stress randomly assigned to administer 300mg CBD or placebo daily for two weeks. Results suggest no significant acute or repeated effects of CBD (verses placebo) on physical symptoms of stress, perceived stress, physical symptoms of anxiety, and state anxiety. There were no significant differences between sexes for any exploratory outcomes. These results suggest there are no acute nor repeated effects of 300mg CBD on stress- and anxiety-related outcomes and no sex differences in this context. Methodological factors may account for the unexpected pattern of results; these are detailed with an eye on informing future, rigorous tests of CBD’s anxiolytic effects
Exploring the Relationship Between Novice Teacher Autonomy and Workplace Engagement, Burnout, and Intentions to Remain in the Teaching Profession
Get PDFAdministrators and experienced faculty frequently disagree on the degree of autonomy that should exist for classroom teachers. Administrators seeking to make adjustments to the levels of autonomy experienced by their teachers may struggle to account for differences between novice teachers and their more seasoned counterparts. This study applies researched-based models for exploring the relationship between various job resources and demands and their ultimate impact on workers. A mixed method designed employing a survey across three large school districts in Arkansas asked participants questions intended to uncover the degree of autonomy novice teachers typically experience, levels of burnout & workplace engagement, and intentions to remain in or exit the teaching profession. Qualitative data intended to identify themes in the experiences of novice teachers related to the aforementioned topics was collected through interviews. The convergence of data indicates that while there may not be a statistically significant relationship between any of the primary variables discussed, there does exist a relationship between the degree of collegiality and trust in a work environment, and novice teacher perceptions and opinions of autonomy
The Role of Arkansas Rural Libraries in Cultural Capital Development
Get PDFThis study explores the role of rural public libraries in Arkansas in fostering cultural capital development among youth. By employing a mixed-methods approach combining quantitative analysis from the 2022 Arkansas Public Library Survey and qualitative insights gathered from semi-structured interviews with rural librarians, the research investigates three core questions: (1) How do rural libraries impact youth cultural capital development? (2) In what ways can these libraries enhance their effectiveness in this area? (3) How do librarians perceive their roles in this developmental process?
Findings indicate that rural libraries serve as pivotal cultural coordinators and community hubs, positively affecting youth engagement through targeted programs, diverse resources, and collaborative efforts. Although traditional library resources, including print materials, remain vital, the study highlights the importance of innovative programming tailored to community interests, including STEM activities and community partnerships. However, challenges such as limited funding, staffing, and perceptions of libraries hinder broader participation among young adults.
Recommendations emphasize the need for improved physical spaces, enhanced staffing—with a focus on hiring Master of Library Science graduates—and active community involvement to redefine librarians\u27 roles as cultural advocates capable of fostering long-term youth engagement. The findings suggest that with strategic investment and strengthened support from county governments, rural libraries could significantly improve their contribution to cultural capital development initiatives, thereby enhancing educational outcomes and economic prospects for the communities they serve
The Evolving Changes in the Regionalization of Export Bans on U.S. Poultry Trade from the 2015 and 2022 Highly Pathogenic Avian Influenza Events
Get PDFAs food security becomes a growing concern worldwide in a trading system dominated by globalization, the safeguarding of a bio-secure food system is at risk. One of the greatest risks to this stability of the system is, highly pathogenic avian influenza or HPAI which is growing in frequency and impact. Therefore, the global response to disease events is evolving into a newfound understanding of disease prevalence across the animal protein industries. This study will analyze the 2022 HPAI event in the United States regarding the regionalization of restrictions imposed by trading partners accounting for trade agreements executed. The examination of potential shifts in trading partnerships, geopolitical relations, trade agreements, and the adaptation or abandonment of regionalized trade restrictions imposed is discussed. Using publicly available panel data coupled with the gravity model, the shifts in trade relations and the effectiveness of regionalization export restrictions against the United States were explored. The conclusions derived from the model show the need for targeted policies as commodities and regions of the world respond differently to the factors of regionalization of bans
Latent Variable Dyadic Regression Models for Predicting Over/Under Bets in Sports Betting
Get PDFThis thesis explores the use of latent factor models to uncover hidden structures in pair wise outcomes derived from Over/Under betting markets in sports betting. Specifically, we implement and evaluate the Eigen model, a latent space model that represents dyadic data using node-specific vectors whose inner product govern edge probabilities. By modeling relationships between teams as adjacency matrices of binary outcomes, we investigate the extent to which the Eigen model captures both homophily, the tendency of similar teams to yield consistent betting results, and stochastic equivalence, where different teams exhibit indistinguishable patterns of Over/Under outcomes. A Bayesian formulation of the model allows for posterior inference on team-level latent traits and model parameters, while poste rior predictive checks are used to assess model fit. We further discuss the potential for such models to detect systematic biases in bookmaker lines, highlighting how latent structures in match-ups may inform profitable betting strategies under market inefficiencies. Simulated and real-world betting data are used to illustrate the model’s capabilities and limitations
Automation of Vulnerability and Patch Management: Information Extraction, Association, and Optimization
Get PDFVulnerability and patch management is an integral part of a robust cybersecurity program, yet it grows increasingly complex due to the sheer amount of data that must be analyzed. Particularly in Operational Technology (OT) environments, analysis must be done manually because of the lack of automated solutions. Additionally, there are many steps in this process, from the initial discovery of the vulnerability to the implementation of its remediation, and each step in the process requires different data in order to be performed effectively. In this work, we provide approaches and strategies to assist operators in industrial or OT environments throughout the vulnerability management cycle. Security advisories provide key information about mitigation strategies, or actions that can be taken when a patch is unavailable or cannot be installed. Details of these strategies are not shared in public vulnerability databases and must be found manually. We approach this problem by designing a solution to automatically identify that information within vendor security advisories and retrieve it for operator use. We start with an approach that requires domain-specific knowledge of certain frequently-seen reference websites. Next, an approach that can work on an arbitrary website but relies on certain keywords. Finally, an approach that uses Natural Language Processing (NLP) methods and does not require specific knowledge or keywords. Each of these approaches is more general than its predecessor; we demonstrate high accuracy for all approaches. Advisories also often contain details of affected products in non-standard or natural language formats. While this information can be easily understood when read by an operator, the non-standard format acts as a barrier to effective automation. We provide an approach for the first step in this process: identifying vendors in security advisories and mapping them to a standard framework for representing digital assets and software products. We evaluate five established string similarity algorithms, plus one of our own design that combines string similarity and information theory, on the task of mapping vendors to their corresponding entries in the Common Platform Enumeration (CPE) repository. Our results show that our proposed metric outperforms all others. Due to the constraints on time, finances, and personnel for organizations, Large Language Models (LLMs) may seem like attractive opportunities for security operators to speed up information gathering; however, it is still not clear whether LLMs can handle vulnerability management tasks well. To answer this question, we perform an empirical study of LLMs’ ability to provide consistent, accurate information about vulnerabilities in order to guide organizations in their adoption of LLMs. We observe poor performance for all models tested, suggesting that these models are not well-suited to the consistent retrieval of accurate vulnerability information. Finally, once vulnerabilities have been identified and any additional information has been obtained, operators must decide which remediation actions to implement based on their available resources. This already-complex problem becomes even more so when we consider that a vulnerability may have multiple avenues for remediation. We formulate this scenario as two knapsack problems and provide solutions, which we then compare against several existing strategies for vulnerability prioritization seen in real operational environments
Utility of John Deere\u27s See & Spray Ultimate in Southern U.S. Row Crops
Get PDFRow crop producers across the southern United States are observing a boom in technological advancement to improve operational efficiency. Additionally, environmental concerns and reduced profit margins drive the adoption of targeted pesticide application technologies like the John Deere See & Spray. This technology is relatively new, introduced in 2022, and producers need more insight into the shortcomings, potential herbicide savings, and the economic benefits of adoption. Since 2021, research has been conducted in Keiser, AR, with additional collaborative projects in Mississippi, North Carolina, Indiana, and Illinois. Five different objectives were addressed in this research: 1) to determine the impact of residual herbicide application methods in soybean [Glycine max (L.) Merr.]; 2) to evaluate targeted application (TA) performance in cotton (Gossypium hirsutum L.) grown with and without cover crops; 3) to quantify the likelihood of treating weeds with TA while considering the dynamics between weeds, crops, and detection settings; 4) to assess the ecological risk of adopting TA at two extreme detection settings in a three-year soybean system; 5) to quantify the opportunity cost between traditional broadcast applications and targeted sprayer systems. The first objective evaluated either TA of all herbicides (single-tank), TA of postemergence (POST)-active herbicides and broadcasting residual herbicides (dual-tank), and different residual application timings in four site years. Targeted applications provided comparable control to traditional broadcast applications across all weeds, resulting in a 28.4% to 62.4% POST herbicide savings. The second objective evaluated single-tank TA, dual-tank TA, and broadcast applications in fallow, cereal rye (Secale cereale L.), or hairy vetch (Vicia villosa Roth] cover crop systems. Overall, TA with See & Spray could detect weeds in cover crop biomass. By the end of the season, weed control for all species, including Palmer amaranth (Amaranthus palmeri S. Watson), was ≥ 92%. For Objectives 3 and 4, the lowest sensitivity settings reduced the likelihood of treating weeds, with Palmer amaranth \u3c 5cm in height being missed 39% of the time with the lowest setting, which ultimately caused an increase in weed density in subsequent years. For areas treated using the lowest sensitivity each year, the weed density increased from 867 weeds ha-1 in 2022 to 11,300 weeds ha-1 in 2024, while plots containing the high sensitivity and broadcast treatments had 2,336 weeds ha-1 at termination in 2024. For the opportunity cost, nozzles with narrower spray plume angles sprayed less, and average break-even points were determined for a John Deere 412R (Deere & Company, Moline, IL) when purchasing a new or upgrading an existing sprayer. Adopting See & Spray appears to be a viable tool to reduce the environmental loading of herbicides and enhance producer profitability if appropriately utilized. Within a program approach, targeted applications can provide comparable weed control to broadcast applications with a medium or higher sensitivity setting. Reduced weed areas improve targeted herbicide savings and emphasize the importance of highly effective residual herbicide programs and integrated weed management strategies. Lower sensitivity settings increased Palmer amaranth density in subsequent years and should not be utilized for typical herbicide applications. Return on investment will depend heavily on the investment cost, intended herbicide program cost, and weed densities at the time of application