Digital Commons @ Florida Atlantic
Not a member yet
400 research outputs found
Sort by
THE HARDY-LITTLEWOOD INEQUALITIES FOR SOME ALMOST PERIODIC FUNCTIONS
No full textThis work is a continuation of Boryshchak, Myers, and Sagher in [10], extending and refining results in the theory of almost periodic functions. New results are accompanied by proofs and unified methods using tools drawn from harmonic analysis, functional analysis, point set topology, the theory of charges (finitely additive measures), and mathematical logic. The central contribution of this work lies not only in its results, but in its exposition, which presents a coherent framework intended to bridge advanced research and graduate-level study, to further developments in related areas
NODE-ATTRIBUTED SPATIAL GRAPH PARTITIONING: A KULLBACK-LEIBLER DIVERGENCE BASED APPROACH
No full textThis thesis presents a novel approach to node-attributed spatial graph partitioning (NSGP) that addresses limitations in existing homogeneity measures. We introduce a Kullback-Leibler (KL) divergence-based formulation we name KLD that measures attribute similarity by comparing each node’s distribution to its subgraph centroid, providing a computationally efficient method of measuring homogeneity. Integrated within a multilevel graph partitioning framework and Fiduccia-Mattheyses refinement, our approach achieves approximately 90% runtime reduction compared to the baseline Clustering and Local Refinement (CLR) algorithm while maintaining comparable partition quality. Applied to COVID-19 infection data combined with American Community Survey demographics across U.S. FIPS codes, KLD successfully identifies spatially contiguous, demographically homogeneous communities. Experimental results on graphs ranging from hundreds to thousands of nodes demonstrate favorable scaling behavior, making the approach practical for large-scale spatial analysis where both geographic proximity and attribute similarity must be jointly optimized
POST-QUANTUM CRYPTOLOGY: NEW CONSTRUCTIONS AND CRYPTANALYSIS
No full textClassical cryptographic schemes, which are based on the hardness of factorization and discrete logarithmic problems, can be efficiently solved by the Shor algorithm on a quantum computer. This motivated the 2016 National Institute of Standards and Technology (NIST) call to identify efficient and secure cryptographic schemes that are resilient to potential attacks from both classical and quantum adversaries, a field referred to as post-quantum cryptography. In this work, we focus on designing efficient post-quantum cryptographic primitives based on code-based and lattice-based assumptions, and we analyze their underlying hardness using quantum cryptanalysis. On the construction side, we propose a ring signature scheme and an identity-based signature scheme based on the Code Equivalence Problem, leveraging the LESS identification scheme and the Calamari-Falafl framework. The proposed ring signature achieves small public keys (11.57 kB), and its signature size grows logarithmically with the number of users in the ring, outperforming existing code-based solutions while remaining competitive with other post-quantum schemes, particularly for large ring sizes. Furthermore, we introduce cryptographic schemes based on the Module NTRU problem, a generalization of the NTRU problem that provides better flexibility in parameter selection. Building on this, we design compact encryption schemes that achieve a low decryption failure rate, with a proposed parameter set offering the smallest ciphertext size among NIST Level 3 security schemes. Additionally, we present the design of signature schemes, one of which achieves the smallest provably secure signature size in the Quantum Random Oracle Model (QROM). On the cryptanalysis side, we present a concrete quantum resource estimation for lattice enumeration based on Montanaro’s algorithm, together with a detailed implementation in the quantum circuit model, and show how to optimize the depth of the circuit through parallelized design components. The second contribution is a quasi-polynomial-time algorithm for the Extrapolated Dihedral Coset Problem (EDCP) over power-of-two moduli. Although our results on EDCP do not compromise the security of LWE with standard parameters, they offer insight into the complexity of LWE
ASSESSING THE HURRICANE INLAND WIND EXPOSURE ALONG THE U.S. COASTAL STATES
No full textLandfalling tropical cyclones (TCs) are among the most destructive U.S. natural hazards, and recent storms show that damaging winds and flooding often extend far inland. This study examines inland TC wind exposure across coastal states from Texas to North Carolina using HURDAT2 data (1900–2024). Following Zhu et al. (2023), an exponential decay interpolation model and an Accumulated Cyclone Energy (ACE) framework were applied to capture continuous variations in intensity and duration. Results show that TC winds can persist hundreds of kilometers inland, with States experiencing ACE levels comparable to some coastal counties. Incorporating storm structure through the radius of maximum wind (RMW) reveals that stronger cyclones are generally more compact, while weaker storms exhibit broader variability. These findings demonstrate that inland regions are more exposed to TC wind energy than previously recognized and improve understanding of inland wind decay and spatial hazard distribution
ASSESSING TROPHIC PATHWAYS IN THE SOUTHERN INDIAN RIVER LAGOON USING FATTY ACID MARKERS
Full text linkThe Indian River Lagoon (IRL) comprises over 200 Km of the Florida’s east coast and is a bar-built estuary making up one of the most biodiverse estuarine environments in North America. Toxins produced by increasingly frequent harmful algae blooms in the IRL are found throughout the food web, from zooplankton to fish, to apex predators. The impact of species characterized as harmful algae can be traced using fatty acid markers, as phytoplankton and algae that make up HABs produce characteristic poly-unsaturated fatty acids (PUFA), which have conservative ratios between trophic levels. This study measured the fatty acid composition of two trophic levels sampled from five locations in the Southern IRL between March 2023 and March 2024. Ratios of fatty acids EPA/DHA, C16/C18 fatty acids, PUFA/SFA, and C16:1/C16 fatty acids showed shifts in the diets of zooplankton from dinoflagellate dominant to diatom dominant, and these ratios were conserved in fish
AI-ASSISTED RADIO FREQUENCY FINGERPRINTING FOR IDENTIFICATION OF USER DEVICES IN 5G NETWORKS
Full text linkAs 5G networks expand, ensuring secure identification and authentication of user devices is critical. This paper explores machine learning-based Radio Frequency Fingerprinting (RFF) to identify and distinguish trusted and rogue devices in 5G networks. We evaluate ResNet, Transformer, and LSTM architectures using channel-isolated (CI) spectrogram and raw IQ slice inputs across varying packet sizes. Results show that ResNet with CI spectrogram inputs achieves the highest device classification accuracy and scalability while mitigating the Next-Day Effect, while the same architecture with IQ Slices is best for rogue device detection. Unlike related works, we emphasize the role of spectrograms in accurately capturing discerning features in 5G signals for scalable RFF applications. These fingerprints strengthen authentication processes against device impersonation at the physical layer of 5G networks. Using real-world 5G datasets from an outdoor wireless network testbed, this study demonstrates the feasibility of AI-driven RFF for secure device authentication
BEST FRIENDS AND POPULAR PEERS AS SOURCES OF INFLUENCE DURING LATE CHILDHOOD AND EARLY ADOLESCENCE
Full text linkAcross the transition into adolescence, changes in the child’s social world give rise to growing peer influence. As adult supervision declines and time with peers increases, children and adolescents become increasingly susceptible to influence from friends and popular peers. Unclear, however, is the relative magnitude of each and whether the scope of their influence varies across different domain behavior. Friends dominate private settings, where behaviors that promote reciprocity are particularly salient. Popular peers dominate public settings, where behaviors that promote hierarchy are particularly salient. The present study concerns the unique hypothesis that peer influence is domain specific, such that best friends influence reciprocity behaviors which are prevalent in private settings and popular peers influence hierarchical behaviors which are prevalent in public settings.
Participants were 780 (386 girls, 394 boys) Lithuanian students (5th-8th grade; M=12.29 years old) from 29 classes in three middle schools. Students nominated and rank ordered their best friends. Participants also completed a standard peer nomination inventory consisting of rosters with the names of all students in the class. Students nominated classmates who best fit the following descriptions: Accepted, aggressive, popular, prosocial, rejected, and unpopular. Additionally, participants completed self-report items describing emotional problems, problem behaviors, physical activity, social media use, and weight concerns. Popularity norms were calculated separately for each socioemotional behavior.
Support emerged for the hypothesis that best friend (top ranked nominated friend) and popularity norm influence is domain specific. Best friends influenced emotional problems (a reciprocity domain behavior), as well as problem behavior and prosocial behavior (cross-domain behaviors), whereas popularity norms influenced social media use and (among older adolescents only) weight concerns (a hierarchical domain behavior). Contrary to hypotheses, there was no evidence of popular peer (first nominated popular classmate) influence over any behavior.
Best friends and popularity norms represent unique sources of influence during childhood and early adolescence. Best friends are particularly influential over internalizing symptoms, as well as prosocial and problem behaviors; popularity norms are particularly influential over social media use and (for older adolescents) weight concerns. The findings underscore the importance of assessing influence across sources simultaneously and recognizing that patterns of influence vary as a function of domain
THE IMPACT OF COLLEGE AND CAREER STUDENT SUCCESS CURRICULUM AND XELLO ON HIGH SCHOOL STUDENTS’ SELF-EFFICACY, ACADEMIC MOTIVATION, AND ACADEMIC PERFORMANCE
Full text linkThe study aimed to determine whether the school counselor-led CCSS program (Brigman & Villares, 2015), a curriculum designed to foster cognitive and social-emotional skills, combined with the technological college and career exploration platform, Xello (2019), enhances participating students’ self-efficacy, academic motivation, and academic performance. The CCSS and Xello program consisted of five 45-minute lessons delivered in the classroom by school counselors once a week for five weeks. A quasi-experimental design was utilized with 115 high school students from Southeast Florida assigned to either a treatment group (n = 71) or a control group (n = 44). Pretest and posttest measures were administered to assess changes in self-efficacy, academic motivation, and academic performance over one semester. Results indicated no significant differences between the treatment and control groups, suggesting that the CCSS and Xello intervention did not uniquely contribute to self-efficacy, motivation, or GPA improvements. However, within-group improvements in self-efficacy and academic motivation were observed over time, suggesting that natural academic development or external factors may have contributed to these changes. Despite the lack of significant findings, this study contributes to the growing body of research on school-based CCR interventions as it highlights the ongoing need for scalable, effective support programs to aid students in their transition to postsecondary education
WHISPERS FROM THE ISLANDS PRESERVING AND REINTERPRETING CARIBBEAN MYTHS THROUGH AUGMENTED NARRATIVES
Full text linkThis thesis explores the role of design as a contemporary medium for preserving, reinterpreting, and disseminating Caribbean cultural myths through visual storytelling. Focusing on using Augmented Reality (AR) and paper sculpture, this research examines how immersive technologies can bridge the gap between oral traditions and digital innovation to safeguard intangible heritage. Centering Caribbean folklore, specifically myths such as the Chickcharney, the Lusca, and the La Diablesse, this work analyzes these stories\u27 historical, symbolic, and social functions within Caribbean identity formation and collective memories.
Drawing on interdisciplinary frameworks from mythology studies, design history, and visual culture, the project investigates how AR can amplify cultural narratives, making them accessible to contemporary and diasporic audiences.
By intertwining fragile, tactile materials with interactive digital media, the project examines colonial modes of preservation. It proposes a design-based methodology that honors folklore\u27s performative and evolving nature. Through transmedia storytelling, this thesis advocates for design as a powerful tool of cultural resilience that commemorates the past and reimagines it for the future