41656 research outputs found
Sort by
Synthesis and Characterization of Magnetic Nanoparticles to Study Effective Magnetic Anisotropy for Biomedical and Catalytic Applications
No full textThis dissertation focuses on understanding how to tune the magnetic properties of nanoparticles through controlling the effective magnetic anisotropy (Keff), which is a key variable in determining a nanoparticle’s Néel relaxation time, which will dictate its magnetic behavior in various applications. In this work, magnetocrystalline anisotropy is tuned by synthesizing tri-metallic substituted ferrite (Fe3-x-yMnxCoyO4) nanoparticles with specific metallic compositions that were informed by computer simulations using density functional theory (DFT) to target magnetocrystalline anisotropy values. A drip synthesis was used to control the size and composition of the tri-metallic ferrites, which were revealed to be monodisperse and compositionally mixed by x-ray diffraction (XRD), high-resolution transmission electron microscopy (HRTEM), and inductively coupled plasma-optical emission spectroscopy (ICP-OES). Both calorimetric and AC magnetometric specific absorption rate (SAR) measurements suggested that the nanoparticle’s heating capabilities were influenced by their metallic compositions. The tri-metallic nanoparticles were then measured using a physical property measurement system (PPMS) to experimentally measure their Keff values. As the tri-metallic nanoparticles increased in manganese content, the Keff values ranged from 80,000 J/m3 to 115,000 J/m3, showing an opposite trend that was predicted through the DFT simulations and illustrating magnetocrystalline anisotropy can be altered via metallic ion substitution in the nanoparticles’ crystalline lattice. A targeted higher cobalt content tri-metallic nanoparticle series was also synthesized, with a Keff value of 153,162 J/m3. To synthesize a multi-modal nanoparticle for biomedical applications, nanoclusters with a distinct shape resembling flowers were synthesized. The iron oxide nanoflowers (IONFs) were also doped with gadolinium (Gd-IONFs) to improve their magnetic heating, and their magnetic resonance imaging (MRI) capabilities. The nanoflowers showed a large reduction in their Ms when doped with gadolinium. Low temperature relaxation features below 50 K were measured, suggesting polydispersity in the crystallite sizes of the nanoflowers, suggesting that the magnetic properties of the nanoflowers are dictated by the size of the individual nanoparticles that comprise it. To study the effects of dipole interactions on Keff, spherical iron oxide nanoparticles that were 10 nm in diameter were synthesized and coated in silica shells with various shell thicknesses using a reverse microemulsion method. The thicknesses were determined using HRTEM. The Keff values of the nanoparticles at each shell thickness were then determined through finding what temperature that the maximum value for the imaginary susceptibility occurs at different frequencies. Nanoparticles with high heating efficiency were surface functionalized with a polyacrylic acid – polyethylene oxide (PAA-PEO) copolymer with a glycan known as Asialo GM2 ganglioside oligosaccharide (aGM2) attached to the end of the chains using click chemistry, which has been shown to bind to Neisseria Gonorrhoeae. Each modification step for the polymer was characterized using nuclear magnetic resonance (NMR). The functionalized nanoparticles were characterized using Fourier transform infrared spectroscopy (FT-IR), dynamic light scattering (DLS), and zeta potential experiments. They were found to have the glycan bound to their polymer shells, which allows them to be used in experiments for the selective killing of Neisseria Gonorrhoeae through magnetic induction heating
Robustness Investigation, Detection, and Defense of Deep learning Models against False Data Injection
No full textThis dissertation addresses the critical challenge of adversarial robustness in deep learning systems, focusing on two fundamental domains: time-series prediction and object detection. As these AI systems become increasingly deployed in safety-critical applications from power grid management to autonomous vehicles their vulnerability to adversarial attacks poses significant risks to infrastructure and human safety.
The first contribution introduces a novel stealthy black-box False Data Injection (FDI) attack specifically designed for quasi-periodic time-series data. Unlike existing attacks that produce easily detectable anomalies, our method generates adversarial perturbations that preserve the underlying periodicity and statistical properties of the data, effectively bypassing traditional anomaly detection mechanisms. We rigorously evaluate LSTM models\u27 vulnerability to this attack and develop real-time detection and mitigation strategies that maintain model accuracy on clean data while providing robust defense against sophisticated attacks.
The second contribution presents a comprehensive evaluation of Detection Transformer (DETR) robustness against adversarial attacks. We extend three classic adversarial attacks FGSM, PGD, and C\&W to the object detection domain and conduct extensive experiments on MS COCO and KITTI datasets. Our findings reveal that DETR models exhibit significant vulnerabilities, with PGD attacks reducing average precision to as low as 0.023. We investigate both intra-network and cross-network transferability, discovering that attacks generated on complex models transfer more effectively to simpler architectures. Analysis of self-attention mechanisms demonstrates that adversarial perturbations successfully disrupt DETR\u27s attention patterns, contradicting expectations of inherent transformer robustness.
The third contribution develops a comprehensive framework for adversarial detection in object detection transformers using data manifold theory. Through rigorous statistical analysis, we demonstrate that DETR\u27s logit distributions violate Gaussian assumptions, establishing Kernel Density Estimation (KDE) as superior to parametric methods for modeling class-conditional distributions. We evolve from a fixed-threshold KDE detector achieving 75-92\% detection rates to a sophisticated machine learning framework that reformulates adversarial detection as supervised binary classification. Our optimized logistic regression classifier, leveraging engineered features combining raw logits and KDE-derived statistics, achieves remarkable performance: 95.9\% accuracy, 94.9\% F1-score, and 96.6\% AUC-ROC across combined attack scenarios, while maintaining computational efficiency for real-time deployment.
Collectively, this dissertation advances the state-of-the-art in adversarial robustness by: (1) exposing vulnerabilities in both temporal and spatial deep learning models through novel attack methodologies, (2) providing comprehensive empirical evidence of transformer-based models\u27 susceptibility to adversarial manipulation, and (3) developing practical, efficient defense mechanisms that balance security with operational performance. These contributions establish a foundation for more secure and reliable deep learning systems in adversarially challenging real-world environments
Pre-service teacher perceptions of an outdoor learning experience within a science methods course
Full text linkGiven the challenges currently facing early childhood science education, this exploratory study investigates the advantages of outdoor learning as part of a teacher preparation program. A group of 49 pre-service early childhood education teachers participated in a day-long outdoor learning experience embedded within their science methods course. Guided by the theoretical lens of embodied cognition, we employed a case study approach to collect and analyze qualitative survey data, using both categorizing and connecting strategies to explore participants’ experiences. The findings reveal four key themes related to pre-service teacher perceptions, in that outdoor learning: (1) represented an engaging experience, (2) contributed towards knowledge and skill development, (3) built the community of pre-service teachers, and (4) impacted frameworks for future classrooms. This research contributes to the growing literature on experiential learning in teacher education and highlights the importance of providing pre-service teachers with opportunities to engage in authentic, embodied science experiences
Early Adult Outcomes of 4-H Program Alumni: A Propensity-Matched Comparative Study
Full text linkA young person’s engagement in high-quality youth development programs should lead to stronger positive outcomes as a young adult. Theoretical literature advances broad indicators that mark success in young adulthood; however, there is a dearth of empirical publications reporting long-term outcomes to support this assumption. We conducted a cross-sectional survey study to report on three long-term outcomes (economic stability, health and well-being, and community involvement) of young adult (aged 19 to 34) alumni of the University of California 4-H Youth Development Program. We compared 4-H alumni outcomes to matched peers using secondary data sources. The 4-H alumni sample demonstrated more positive results than the comparison samples on almost all indicators (except family income). Admittedly, there are many challenges in conducting this type of research. We sought ways to overcome the significant biases inherent in this type of research and encourage future empirical research to grow the literature reporting long-term young adult outcomes experienced by previous participants in youth development programs
Item 1: Syllabus HSSJ 2200: Civic Leadership
Full text linkThis is a course in leadership studies that challenges students to consider the leadership development literature in civic contexts. It is also a service-learning course. While service itself is a form of civic participation, it\u27s important to note that service-learning courses don\u27t always focus on issues of civic identity and engagement as much as one might think. Civic learning is not an inherent outcome of all service-learning experiences. The literature in community engaged scholarship is clear that we do not achieve civic learning outcomes from simply adding ten “service hours” to an otherwise unaltered course. Instead, assignments and classroom discussions need to challenge students to draw connections between their community-based experiences and assigned texts, ultimately drawing conclusions about their civic values. My aim in sharing a service-learning course is to provide an example that can be adapted to other subject-matter areas
Asignación Dos: Representación en Redes Sociales de los Representación en Redes Sociales de los Partidos Políticos de Puerto Rico
No full textEl propósito de esta asignación es que los estudiantes analicen cómo los principales partidos políticos de Puerto Rico se representan a sí mismos en las redes sociales. Para ello, re-alizarán un análisis de contenido de las publicaciones de Facebook de cada partido. En esta tarea, los estudiantes desarrollarán habilidades de pensamiento sintético y analítico y apren-derán técnicas de análisis de contenido
Item 6: Service-learning Assignment
Full text linkMany of the stories included here take place outside of the United States. India, South Africa,Senegal, and Guatemala feature prominently while China, Cameroon, and Native lands are also referenced. Other stories take us to places that might be unfamiliar or forgotten: Middle school buses, rural back roads, New York subways, deep South barbershops, sunny Miami beaches, andlonely snowy roads on a winter night. Stories help us to not just enter into the lives of people but also their environments.
This service-learning project is designed not just to help students think about engaging in philanthropic work in a group context but also to consider the lives of civic leaders and civic educators in Africa. Students learn the local challenges faced by these talented leaders and work to help them. Along the way, they are charged with learning a bit about that cultural context as they write a short reflection on the food, dance, dress, or famous tourist site of their assigned country in Africa. As a practical note: While this assignment utilizes an existing relationship with Mandela Washington Fellows, nearly every university has similar relationships with international alumni, visiting academics,or grant-funded global programs
Item 1: Syllabus Introduction to U.S. Government and Politics
No full textThis is an adaptable syllabus for an introductory US Government and Politics course. It provides learning outcomes, an outline for seven 2This is an adaptable syllabus for an introductory US Government and Politics course. It provides learning outcomes, an outline for seven 2-week modules, and the spacing of in-class activities and other assignment
Item 4: Activity 2: Electoral College
Full text linkThis is an activity designed to help students understand how the winner-takes-all process works in the Electoral College and how it can lead to the popular vote winner losing the Electoral College vote