12466 research outputs found
Sort by
Modeling Novel Spin Current in 2D Topological Insulator Materials
Full text linkTopological materials are unique states of quantum matter whose inherent non-trivial properties can give rise to exotic charge and spin transport. Over the past decade, the field of condensed matter physics has extensively studied various features of topological materials. In particular, theorists and instrumentalists have examined the protected surface states of topological insulators (TI) for applications in spintronics. Transport via spin-polarization is an attractive alternative to charge transfer due to less dissipation and ease of tuning. One particularly studied transport mechanism in topological insulators is the induction of spin-polarized charge currents via infrared or terahertz radiation, also known as the Photogalvanic effect (PGE). Most of the experiments have been done in 3D topological insulators. In these experiments, oscillating currents are generated in response to linear and circularly polarized light, which can be tuned by shifting polarization and the angle of incidence. A large body of theoretical literature has also been developed to describe the observed PGE as charge currents in materials with broken inversion symmetry as a second-order non-linear response. In recent years, more attention has been given to 2D topological insulators as an alternative to 3D materials due to their unique transport properties and small band gap. In this thesis, we propose an ultra-thin film 2D TI model and examine its topologically preserved features. Our model resembles a quantum well or atomically thin layers made from 3D TI materials with inversion and time reversal symmetries. The top and bottom surfaces are coupled, producing a band gap of . We calculate the spectrum of the electronic states and derive the equations of motion for charge and spin currents and the trajectory on the coupled surfaces. Our derivations predict spin-momentum locking of surface electrons, which leads to a helical spin trajectory with a precession frequency proportional to the magnitude of electron momentum. Given that the collisions in this type of system are mostly elastic, the spin precession frequency is conserved unless impurity perturbationssignificantly change the potential energy of the electron. We further examine particle-hole transitions in our 2D TI model in response to polarized radiation and characterize the PGE using the Fermi-Dirac distribution function and first-order perturbation theory. Our results show that linear and circular PGE are observable as pure spin currents in our model by projecting currents to a single surface. Numerically, our plots mimic the observed experimental data when recorded from 3D TI surface state measurements. Our results also show the dependence of the spin current magnitude on polarization and the angle of incidence of the incoming light. We, therefore, introduce a novel 2D TI model for tunable spin-polarized transport in response to infrared radiation that can be used for spintronics applications
Ranking Sustainability of Global Fishery Systems Through the Fishery Performance Indicators: A Multivariate Analysis of Triple Bottom Line Outcomes
Full text linkFisheries are complex social-ecological systems requiring context-dependent management approaches to achieve sustainability. While traditional fisheries management strategies have historically focused primarily on the ecological status of fish stocks, it is becoming more commonly recognized that economic and social prosperity must also be evaluated and managed. The Fishery Performance Indicators (FPIs) evaluate fisheries for three pillars of sustainability, producing Triple Bottom Line (TBL) outcome metrics in economic, community, and ecological sustainability. This broadly applicable tool utilizes expert surveys to address the gaps in socio-economic data, assessing the three elements of sustainability separately and including assessments on data deficient fisheries, allowing for comparisons to be made among fisheries globally. This thesis research applied multivariate data analysis techniques to an FPI dataset of 153 fisheries provided by the University of Florida revealing insights on TBL outcomes on a global scale. I examined the cross-sectional links between 17 enabling factors and the TBL sustainability scores, investigating the differences between developing and developed countries to understand their limitations and identify areas of management that need improvement. Since the 1990s, fisheries performance has improved in developed countries with intensive management; however, developing countries face additional socio-economic challenges due to the nature of the national economy, quality of infrastructure, strength of governance, general environmental health, and strength of fishing access rights. Cooperative management strategies promoting stakeholder involvement are needed to address social injustices such as gender inequality in developing countries that may be hindering the development of communities and fisheries sustainability
INTEGRATIVE MULTI-OMICS INVESTIGATION OF BIOMARKERS FOR AGE-RELATED MACULAR DEGENERATION
No full textThis dissertation explores the complex etiology of age-related macular degeneration (AMD) through integrative multi-omics analyses using data from the Age-Related Eye Disease Studies (AREDS and AREDS2), and the Framingham Offspring Eye Study (FOES). The goal is to identify molecular biomarkers and bio-pathways most strongly associated with AMD, particularly the late stages. Metabolomics data from AREDS and AREDS2 revealed key metabolic changes associated with late AMD in both risk and protective directions, identifying sphingolipids and glycerophosphocholines, which may serve as targets for early detection and intervention. Subsequent proteomic analysis from AREDS delineated protein changes, which emphasized the roles of inflammation, complement cascade, and lipid metabolism in AMD’s progression to late AMD. The final segment of this research integrated these findings with data from FOES, using joint pathway enrichment testing and pathway clustering to find the strongest interactions between metabolic and protein pathways and the shared biomarkers across cohorts. This comprehensive analysis highlights the significant overlap in immune and inflammatory responses across different data sets in any-AMD and late AMD states. By finding significant commonalities from three large-scale cohorts, this work enhances our understanding of AMD’s underlying mechanisms and highlights the potential for high throughput, omics-based medicine that could transform AMD management and treatment
Complexity in Animals Color Pattern: Estimating Sources of “Non-biological” Variation in the Study of Color Pattern
Full text linkColor pattern plays a crucial role in various aspects of an organism's biology, including camouflage, mating, and communication. Despite its significance, methods to capture, quantify, and study color pattern variation are often lacking, particularly for complex patterns that defy simple categorization. In this study, we examine existing methods for identifying and capturing animal color patterns and compare them with a new algorithm developed for this purpose. We also created codes to extract data from digital images and measure 19 different pattern elements, aiming to capture the complexity of the color pattern of the Eastern box turtle. One key element we focused on was the symmetry of the pattern on the turtles' scutes, which is of interest in developmental and evolutionary studies. While much work on pattern symmetry relies on theoretical measures rather than empirical data due to quantification challenges, we aimed to address this gap. We chose the Eastern box turtle as our study species due to its accessibility for field photography, clear color pattern against a dark background, rigid shell structure, and lack of extensive pattern research. Our analysis included 55 individuals sampled from both the field and a museum. We evaluated various factors, such as the repeatability and accuracy of the pattern identification algorithm, image acquisition methods, lighting conditions, and animal shape, to understand their influence on pattern variation. It was crucial to account for these factors, as they contribute to non-biological variation and can affect data quality. We also incorporated a citizen science approach to assess color pattern complexity and determine which pattern elements best describe this complexity. Our findings suggest that the developed algorithm is robust and highly repeatable. Elements based on ratios, contrast, or coloration measures were more robust than those relying on the accuracy of the number of pattern objects. Additionally, angle influenced some variation, likely due to technical error. We observed some correlation between certain pattern elements and perceived complexity based on human interpretation of the color pattern. In conclusion, our study represents a significant advancement in the field by providing codes and a workflow for capturing and quantifying color pattern variation in organisms with complex patterns. It also helps differentiate between biological and technical variation in color patterns, improving our understanding of these intricate biological features
Exploring Variability Between Crime Hot Spots Through the Perceptions of Residents
Full text linkResearch on crime and place has made great strides in recent decades. A substantial amount of literature demonstrates that crime concentrates at the micro-geographic level across both place and time. These micro-places with high levels of crime, typically referred to as crime hot spots, have had a key influence on crime policy and theory. Indeed, there is strong evidence that targeting crime hot spots can be effective in reducing crime. Additionally, scholars are finding that crime hot spots are not necessarily just hot spots of crime but also, for example, hot spots of health problems and negative community sentiments. However, the majority of research focused on crime hot spots either examine the phenomenon in relation to “cold” spots, or places that experience little to no crime, or focus on crime outcomes within hot spots. Our knowledge regarding variability between crime hot spots themselves is much more limited. This dissertation by articles aims to address this literature gap by exploring variability between crime hot spots through the perceptions of those living in them. In Chapter 2, I use responses from a random sample of residents living in 50 different crime hot spots in Phoenix, Arizona, to assess their level of support for police in their communities. Specifically, this study first asks if residents want more, the same amount, or less policing. Second, I ask whether this attitude is dependent on certain individual and contextual factors, including demographics, perceptions of police performance, and block context. In Chapter 3, I utilize data from a multi-site randomized control trial to assess variability in collective efficacy at the micro-geographic level across four unique U.S. cities. Here, I ask whether collective efficacy in crime hot spots varies across street segments and across cities. If so, what predicts this variability, and do these influences vary by city? Multilevel models are employed to explore these issues. Overall, this dissertation points to the importance of more closely examining crime hot spots themselves, particularly through the perspectives of those who live in places experiencing disproportionate amounts of crime. This dissertation suggests that during a time of heightened friction between police and civilians, police leaders must develop strategies that reduce and prevent crime in these crime hot spots without alienating residents or contributing to community deterioration
A Stylistic Analysis of Gong Xiaoting's Selected Piano Works
Full text linkGong Xiaoting (b.1970) is a Chinese composer, theorist, and music educator, who serves as a professor at the Central Conservatory of Music in Beijing of China. As an active composer, she has created a large number of excellent compositions and made a huge contribution for the development of composition in her field. Gong’s musical style is diverse and full of innovative thoughts based on her experiences studying in France at her early age. Gong perfectly combines Western impressionistic musical styles with Chinese musical style, In this dissertation, I mainly focused on analyzing her compositional style from her two representatives pieces: (a) Five Paintings in Light Color, and (b) a fugue titled Yili Dance. The first Chapter will provide (a) a brief biographical representation, (b) a brief historical background of Chinese piano music and (c) the need for the study. In the chapter two, I mainly focus on the introduction of Chinese music theory, including pentatonic scales, hexatonic modes, and Heptatonic modes. Meanwhile, I will summarize Gong Xiaoting’s major piano works in the chapter three and let people know what kinds of genres she composed by annotating some typical piano works to further understand her musical style. I mainly summarize her compositions from several perspectives including feminism, fusion style, and nationalism. Chapters four and five includes a thorough analysis of Five Paintings in Light Color and Yili Dance along with performance suggestions for the educators and performers
Ground-based Light Curve Follow-up Validation observations of TESS object of interest TOI 3521.01
Full text link“TESS Mission was launched on April 18th, 2018. It was designed to find the exoplanets by using the transit method. Our goal is to confirm TOI-3521.01 by analyzing the light curve and its NEB check. We use AIJ to do preparation, such as reducing data, plate solving, generating measurement tables, and creating light curves. Then, we examine light curves of our interest by looking for a similar depth and duration as predicted. Our data leads us to be inconclusive, which means that we cannot make sure there is an exoplanet orbit around our host star, because we didn’t rule out false positives.
MACHINE LEARNING AND FEDERATED LEARNING-GUIDED DEFENSE FOR CYBER-SECURITY
Full text linkIn response to the ever-evolving landscape of cybersecurity threats, this dissertationpresents a comprehensive framework that integrates insights from diverse research endeavors, spanning traditional computing paradigms, Internet-of-Things (IoT) ecosystems, energy-constrained devices, and innovative energy harvesting methodologies. The framework is designed to address critical challenges in malware analysis, detection, and defense while simultaneously optimizing energy usage in resource-constrained environments. Beginning with a focus on malware analysis and detection, the framework adopts a dualpronged strategy to efficiently detect traditional and stealthy malware variants. Leveraging machine learning classifiers and recurrent neural networks (RNNs) on microarchitectural traces and localized feature extraction from binary images, the framework achieves impressive detection accuracies of up to 94% for traditional malware and nearly 90% for stealthy variants employing code relocation obfuscation techniques. Transitioning into the IoT domain, where security vulnerabilities are exacerbated by interconnected smart devices, the framework introduces a collaborative machine learningbased malware detection framework. Incorporating performance-aware precision-scaled federated learning (FL) and the active defense strategy RAPID (Robust and Active Protection with Intelligent Defense), the framework achieves an average accuracy of 94% with minimal communication overhead. These results outperform existing techniques by 19% while ensuring data security and privacy. Recognizing the unique challenges posed by decentralized training data in FL and the energy constraints of IoT devices, the framework integrates energy harvesting methodologies. A novel RAFeL (Robust and Active Federated Learning) framework is introduced, combining robust defense mechanisms and performance-aware bit-wise encoding. This approach achieves higher compression rates and minimizes communication costs while maintaining model accuracy within a narrow range of 3% to 10%. Further extending the focus to energy-constrained devices, the framework proposes a methodology integrating kinetic energy harvesting with machine learning tasks. By optimizing energy usage through checkpointing techniques and Energy-aware Early Exit Neural Networks, the framework effectively reduces reliance on built-in batteries. This approach enhances device longevity while maintaining computational efficiency, making it well-suited for applications in IoT and edge computing environments. Complementing these advancements is IRON-DOME, an integrated defense system tailored for IoT networks. By combining image-based malware detection, dynamic behavior analysis using Hardware Performance Counter (HPC) values, and network packet data analysis, IRON-DOME achieves a runtime malware detection accuracy of 93% with notable resource and power efficiency gains. It consumes 30% fewer resources and 40% less power than state-of-the-art defense techniques, ensuring robust protection against evolving cyber threats. Through these integrated research endeavors, this dissertation contributes to a holistic understanding of contemporary cybersecurity challenges and presents innovative strategies to fortify security in diverse computing environments. The framework’s scalability, adaptability, and efficiency make it a valuable asset for mitigating cybersecurity risks while optimizing resource utilization in the digital era
Predicting Alzheimer’s Disease from miRNA Sequence and Expression Data with Machine Learning
Full text linkApproximately 6.5 million people, most of whom are 65 years of age and older, have been diagnosed with Alzheimer's Disease (AD) in the United States. Diagnosing AD has notoriously been difficult because disease progression can occur before the onset of cognitive impairment, and the physiological changes in AD brains are largely only observable in post-mortem studies. AD screening has been bolstered by novel biomarkers, including expression profiles of exosomal and circulating miRNAs. Although relatively new to biological studies, these miRNAs have become a focal point due to their widespread availability in bodily fluids and potential use in disease diagnostics. The purpose of our study was to investigate the utility of machine learning (ML) to predict AD-associated outcomes with miRNA sequence and expression data. Machine learning was performed leveraging the Orange Data Mining platform, which allowed us to quickly prototype various machine learning models and assess their performance numerically and graphically. To utilize miRNA sequence data, we employed a k-mer bag of words model to quantify subsequences within miRNAs and predict if miRNAs are involved in AD pathways. We found that a random forest model provides the best predictions with an accuracy of 0.772 and an area under the receiver operating characteristic (AUROC) of 0.813. Interestingly, out all k-mers, we found that those rich in purines are the most predictive of miRNA association with AD. As a second modelling effort, we analyzed a previously published dataset [Ludwig et al. (2019) Machine Learning to Detect Alzheimer’s Disease from Circulating Non-Coding RNAs Genom. Proteom. Bioinform. 17(4): 430-440] that measured miRNA expression in AD and healthy patients. A random forest model produced an accuracy of 0.786 and AUROC of 0.862 approximately reproducing the published results. We explored if the likelihood for miRNAs to be associated with AD-related pathways can be used as additional selection criteria for miRNA expression profile analyses and discuss the broader applications of our machine learning models in AD diagnostics. Ultimately, we believe our machine learning models will be useful to determine for new miRNA sequences if they are likely to be involved in AD and to pre-select miRNAs as biomarkers for expression profile analysis, which could be used as a diagnostic tool
George Mason, Luther Martin, and the Origins of State Power in American Constitutionalism
No full textGeorge Mason of Virginia and Luther Martin of Maryland became leading figures in the debate over the ratification of the United States Constitution. Their contributions to the emerging nation are due for a reassessment. This dissertation makes the case for Mason and Martin as not just two among many Anti-Federalists but as figures who were among the most important Anti-Federalist leaders of the time. Both men attended the Philadelphia Convention and made significant contributions in shaping the Constitution. Afterward, their decision to oppose the document had a powerful impact on the public debate. Martin and Mason used the power of the press to voice their views and extend their influence far beyond their individual states. Significantly, their publications, published under their own names, were reprinted more times than other Anti-Federalist works during the ratification debate. While they did not prevail in the ratification debate, their ideas concerning individual liberty and the centrality of state power persisted long after the Founding Era. Besides their shared political views, Mason and Martin also shared other important traits: both were slaveholders and both came the country’s Chesapeake region. Unlike northern states, Virginia and Maryland did not move to abolish slavery during or after the American Revolution. Unlike the states of the Deep South, in their states’ slavery was in decline. Their distinctive regional origins and background shaped their particular version of Anti-Federalism. Both men supported preserving state power but both spoke out against either slavery or the slave trade. Neither Mason and Martin portrayed state power as a means of defending the institution of slavery. In later decades, southerners would appropriate their ideas and mold them into a states’ rights argument that would be used to preserve slavery. Despite their inability to derail the Constitution, both Mason and Martin also made other substantive contributions to the development of American constitutionalism. At the Philadelphia Convention, Martin introduced the Constitution’s Supremacy clause, arguably the most important provision in establishing the federal government’s authority over the states. During the ratification debate, Mason’s Objections to the Constitution became a rallying point for those who desired the addition of amendments to the document. By making individual and state liberties a central issue, Mason and other Anti-Federalists put pressure on the First Congress to pass a Bill of Rights. Long after ratification, Luther Martin continued to press the case for preserving state power. He represented his home state in the landmark 1819 Supreme Court case of McCulloch v. Maryland. Although Martin lost, McCulloch addressed the extent of the federal government’s power in light of the Constitution’s Tenth Amendment, Supremacy Clause, and Necessary and Proper Clause. George Mason and Luther Martin have had a long-lasting and hitherto understudied impact on the development of American constitutional thinking. Through their efforts, the plain language of the Tenth Amendment has kept the hope alive of state power in face of the increased power of the federal government. During the antebellum era, southerners distorted their views about state power to serve the interests of slavery. By the twentieth century and beyond, supporters of states’ rights would resurrect their views for other purposes. Yet Mason and Martin’s ideas must be considered in the context of their own time in order to understand their present-day significance. In both instances, the effectiveness of their distinctively Chesapeake defense of state power in opposing the Constitution has made an enduring impact on American constitutionalism