Digital Commons Network^™

Pinnlab: An Interactive Dashboard For Teaching Data-Driven Parameter Estimation In Differential Equations Using Physics-Informed Neural Networks, Mohan J. Parthasarathy, Padmanabhan Seshaiyer

CODEE Journal

Undergraduate instruction in ordinary differential equations (ODEs) is typically organized around the forward problem: finding solution trajectories when the governing equation and its parameters are known. In scientific practice, however, inverse problems are often more relevant, requiring unknown parameters to be inferred from noisy observations while assessing whether a proposed model is consistent with the data. We introduce PINNLab, an open-source MATLAB dashboard designed to help undergraduate students explore inverse modeling through physics-informed neural networks (PINNs). PINNLab presents PINNs as a complementary data-driven framework that connects differential equations, optimization, empirical data, and scientific machine learning. The instructional sequence is organized …

Reproducing And Analyzing The “Lost In The Middle” And “The Power Of Noise” Phenomenon In Retrieval-Augmented Generation, Kousik Samanta

Master’s Dissertations

Retrieval-Augmented Generation has become the way to improve Large Language Models. They help with problems like knowledge and hallucinations. Recent studies show that these models still have limitations. One big problem is the “Lost in the Middle” phenomenon. Models can’t access information in the middle of contexts properly. Another counterintuitive observation is the “Power of Noise” paradigm, which suggests adding unrelated documents can actually make the generation better. We know these happen in extractive QA tasks, but we don’t know if they happen in tasks that need complex reasoning. This dissertation looks into how position and noise affect Long-Form Question …

Learning To See Lesions, Not Skin Tone: Counterfactual Multimodal Learning For Fair, Trustworthy, And Text-Free Dermatology Ai, Shivam Jangid

Master’s Dissertations

Recent advances in deep learning have significantly improved the performance of automated skin lesion classification systems, enabling accurate detection of various dermatological conditions from medical images. Despite these achievements, concerns regarding fairness and generalization remain a major challenge for the deployment of such systems in real-world clinical settings. A key factor contributing to these challenges is the presence of bias in training datasets, particularly with respect to skin-tone representation. Most publicly available skin lesion datasets contain a disproportionate number of samples from individuals with lighter skin tones. As a result, deep learning models trained on these datasets often learn representations …

Adaptive Spectral Trust Gate For Physics- Constrained Operator Learning, Soham Chakraborty

Master’s Dissertations

Physics-informed machine learning improves the plausibility, data-efficiency and generalization of surrogate models by injecting prior physical knowledge into the learning process. The current approaches can be broadly divided into two main categories: soft constraints, which add a physics residual to the training loss but guarantee nothing at inference time, and hard constraints, which project the model output onto the constraint set exactly but apply the projection uniformly to every part of the signal — including parts that are dominated by noise, discretization error, or model mismatch, where the idealized physics is not actually trustworthy. This dissertation proposes the Adaptive Spectral …

Deep Reinforcement Learning With Directed Asymmetry And Kolmogorov-Arnold Networks For Dismantling Interdependent Multiplex Networks, Soumyajit Dev

Master’s Dissertations

Identifying the minimum-cost node-removal sequence that fragments a complex network - the network dismantling problem is NP-hard and central to infrastructure resilience. In interdependent multiplex networks, this difficulty is compounded by cascading cross-layer failures. While deep reinforcement learning (DRL) agents utilizing graph neural network (GNN) encoders achieve near-optimal dismantling, current state-of-the-art architectures suffer from two critical limitations. Topologically, existing agents strictly assume undirected edges, rendering them inapplicable to directed systems - such as supply chains or gene regulatory cascades - where failure propagation is fundamentally asymmetric. To resolve this, we propose Disassembling Directed Interdependent Networks (DDIN). DDIN introduces an asymmetric …

Antioxidant Potential Of A Non-Heme Iron Complex: Mitigating H2O2-Induced Cellular Damage, Magy A. Mekhail, David M. Freire, Cameron Bowers, Grant Elam, Nora Del Bosque, Hannah K. Pyle, Sarah K. Dunn, Spencer Lanyon, Timothy J. Hubin, Giridhar Akkaraju, Kayla N. Green

Biology, Chemistry, and Environmental Sciences Faculty Articles and Research

Iron complexes with antioxidant activity have garnered significant interest for both general and medical applications, yet few studies have yielded functional nonheme iron-based mimics for H₂O₂ mitigation in aqueous environments. Here, we investigate water-soluble nonheme iron complexes derived from 12-membered pyridinophanes (^CF₃PyN₃, PyN₃, ^NMe₂PyN₃, and Py₂N₂) to evaluate the impact of ligand scaffolds on H₂O₂ decomposition activity both in vitro and in a cellular model. Speciation and kinetic analyses reveal that both electronic and geometric modulation of the macrocyclic …

Adding 3,4-Methylenedioxy-Α-Cyclohexylaminopropiophenone To A Standard Operating Procedure (Sop) For Identifying Synthetic Cathinones Using Gas Chromatography-Mass Spectrometry, Napassawon Timpratoom

ETDs from 2020-2029

Synthetic cathinones were identified as new psychoactive substances (NPS) in the early warning advisory system by the United Nations Office on Drugs and Crime (UNODC). Synthetic cathinones have been commonly referred to as bath salts, which offered “legal highs” because they were not explicitly scheduled. These compounds were detected dur-ing the screening process in casework and required internal validation for confirmation. 3,4-methylenedioxy-α-cyclohexylaminopropiophenone was the target analyte for this study. It was alternatively named N-cyclohexyl methylone and “cyputylone." The aim of the study was to validate 3,4-methylenedioxy-α-cyclohexylaminopropiophenone into a Standard Operating Procedure (SOP) for identifying synthetic cathinones using gas chromatography/mass spectrometry …

New Studies In Lattice-Based Cryptography, Quantum Algorithms, And Privacy-Preserving Computation, Hansraj Jangir

Electronic Theses and Dissertations 2020 - Present

Classical public-key cryptographic schemes primarily rely on the presumed hardness of the integer factorization and discrete logarithm problems. However, Shor’s algorithm demonstrates that both problems can be solved efficiently on a sufficiently powerful quantum computer. This breakthrough motivated the search for quantum-resistant hard assumptions. The assumptions based on lattices are one of the primary candidates in post-quantum cryptography due to their strong theoretical foundations and well-established hardness against both classical and quantum attacks. In this thesis, we explore several directions of lattice based cryptology and quantum algorithms.

On the construction side, we propose two compact encryption schemes based on the …

Interseismic Creep Along The Enriquillo–Plantain Garden Fault, Haiti, Estimated From Insar, Rishabh Dutta, Jeremy L. Maurer, Yi Chieh Lee

Geosciences and Geological and Petroleum Engineering Faculty Research & Creative Works

The Enriquillo–Plantain Garden fault zone (EPGFZ) is a major left-lateral strike-slip fault in southern Haiti hosting several recent destructive earthquakes, including the 2021 MW 7.1 Nippes event with primary thrust-slip. To investigate how strain is accommodated in the vicinity of the 2021 event, we analyzed Sentinel-1 interferometric synthetic aperture radar (InSAR) data from 2017 to 2021 using a PS + DS (persistent + distributed scatterers) InSAR time-series approach to overcome decorrelation in this highly vegetated region. We reveal interseismic creep along the EPGF and adjacent Grand'Anse–Sud border faults, with pure strike-slip mechanism and rates up to 9 mm/yr. We suggest …

A System For The Prediction Of Election Results Using Vader And Hybridized Machine Learning Model, Abraham E. Evwiekpaefe, Khadijah Kabir, Georgina N. Obunadike

Tanzania Journal of Science

Integrating different classifiers along with sentiment lexicons like Vader, can enhance the performance of sentiment analysis systems. However, such a hybrid model remains underexplored, particularly in the context of regional elections in developing countries like Nigeria. The aim of this research is to develop a hybrid model that combines three machine learning classifiers and Vader lexicon to possibly achieve a higher accuracy. A case study of the 2023 governorship election in Kogi, Bayelsa and Imo State, Nigeria was examined. Twitter API library was utilized to extracted public and personal tweets using hashtags and keywords related to the target data from …

Robust Real-Time Uav Target Tracking With Onboard Vision-Based Yaw Control, Rylan Malarchick, Jose Castelblanco, Enrique Amaya, Carmen Dimario, Graysen Brinkman, Chirag Kumar, Kiwon Yoon, Sajid Berhane

Beyond: Undergraduate Research Journal

Autonomous tracking of agile unmanned aerial vehicles (UAVs) presents significant challenges for real-time perception and control systems. This work presents AIRHOUND (Autonomous Intelligent Rotorcraft for Hostile Object Unified Navigation and Detection), a UAV platform implementing vision-based yaw tracking through a modular ROS2 software architecture. The system employs YOLOv8 object detection optimized with NVIDIA TensorRT for embedded deployment on an NVIDIA Jetson Orin companion computer. Detected targets are processed through a geometric tracking module that converts pixel coordinates to angular yaw errors using pinhole camera intrinsics, with a proportional controller generating rate-limited yaw commands. These commands are streamed to a PX4 …

Efficiency Of Dairying Under Contrasting Feeding And Management Systems In North America, Israel, Europe, And New Zealand, J.D. J. Scott

IGC Proceedings (1977-2023)

Four different systems of feeding and managing dairy cattle are seen in major milk-producing countries. These are (1) feeding purchased roughages and concentrates to zero-grazed, intensively penned cows; (2) home growing of roughages and most con­centrates for zero-grazed animals; (3) feeding concentrates with conserved pasture and other roughages in winter and reduced concentrates to grazed cattle in spring-summer; and (4) feeding cattle on grazed pasture and conserved pasture throughout the year. The following data on the four systems were obtained from examples in southern California, Wisconsin, England/Wales, and New Zealand. Average fat-corrected milk (FCM)/cow is approximately 7,140, 5,300, 4,670, and …

Spectral Unmixing Using Machine Learning, Debashis Dhar

Master’s Dissertations

Spectral Unmixing is an important field of study nowadays which focuses on gener ating fractional abundance of each pixel into constituent materials .In this thesis we have tried to unmix each pixel into three end members namely glacial lake,debris and others with primarily focusing on glacial lake.We have performed various meth ods of linear spectral unmixing and non linear spectral unmixing. These methods are applied on the collected LandSat Data of east Himalayan terrain .Experimental results demonstrate the effectiveness of the proposed approach in achieving high accuracy and efficiency in glacier lake tracking on LandSat data.

A Switch-Point-Aware Contrastive Approach To Sentiment Analysis Of Hinglish Code-Mixed Text, Prasant Kumar Sahoo

Master’s Dissertations

With the increasing use of social media in non-English-speaking regions, especially in India, people often use Romanized Hindi and English together in their online communication. In a single sentence, they frequently mix Romanized Hindi and English, creating code-mixed text. However, most multilingual transformer models are pre-trained primarily on monolingual data. As a result, NLP systems face challenges when processing code-mixed text, as a single word may be fragmented into meaningless subword pieces, making it difficult for the model to capture its semantic meaning accurately. In this dissertation, we propose a parameter efficient neural architecture consisting of three main components to …

Simultaneous Tumor Delineation And Report Generation From Brain Mr Images, Adish Mallik

Master’s Dissertations

Brain tumor analysis is an important application of medical image computing, where accurate segmentation and interpretation of tumor regions can support diagnosis and treatment planning. However, existing methods often address tumor segmentation and radiology report generation as separate tasks. Moreover, one of the major challenges in report generation tasks from MRI is accurate tumor localization. Most models fail to locate the lobe and hemisphere in which the tumor is located, causing incorrect report generation. In this regard, a unified 3D vision-language framework is proposed for simultaneous brain tumor segmentation and report generation from multi-modal MRI. Given T1, T2, T1C, and …

Case Study On Operational Performance Improvement By Turbocharger Nozzle-Ring Modification In Shipboard Generators, Jong-Hun Im, Mukyeong Kang, Sang-Deuk Lee

Journal of Marine Science and Technology–Taiwan

Owing to increasingly stringent emission regulations imposed by the International Maritime Organization and rising fuel prices, the demand for enhanced energy efficiency and emission control in marine power systems is increasing. This study presents a case study in which the performance of an aging auxiliary generator is restored by modifying the turbocharger nozzle ring. The vessel, a 12-year-old crude-oil tanker, experienced repeated operational issues with its No. 3 generator. This included elevated exhaust gas temperatures, reduced intake air pressure, and frequent turbocharger surges. To address these challenges, the nozzle-ring blade angle was reduced by approximately 2° using spare components, while …

Towards Sustainable Marine Education: Assessing High School Students’ Understanding Of Conservation Concepts, Cheng-Chieh Chang, Tsuyoshi Sasaki, Ya-Yin Liao, Ling-Ya Hung, Liang-Ting Tsai

Journal of Marine Science and Technology–Taiwan

This study assesses the concepts of marine resource conservation and sustainability among Taiwanese high school students by developing a Chinese version of the Marine Resources Conservation and Sustainability Scale. Grounded in the cognitive, socio-emotional, and behavioral learning objectives outlined in UNESCO’s Education for Sustainable Development Goals (ESDGs), the scale targets students aged 16 to 18 to examine gender differences in marine knowledge and conservation concepts. The scale comprises 31 items across four dimensions: the first section evaluates marine knowledge, while the remaining dimensions utilize a four-point Likert scale to assess cognitive, socio-emotional, and behavioral learning objectives, respectively. The results indicate …

Dynamic Governance Of Transboundary Fisheries Through Deep Learning And High-Frequency Vessel Tracking Data For Climate-Resilient Management: An Application To The East China Sea, Qin Lin, Yaqing Shu, Manel Grifoll, Hongxiang Feng

Journal of Marine Science and Technology–Taiwan

Transboundary fisheries governance is increasingly challenged by jurisdictional conflicts, regulatory asymmetries, and climate-induced stock shifts in fish stocks, especially in contested regions like the East China Sea (ECS). This study capitalizes on China’s high-frequency BeiDou Navigation Satellite System (BDS) data (June 2021–May 2022) to pioneer a CNN-BiLSTM deep learning model to dynamically identify fishing hotspots and evaluate policy effectiveness. By integrating vessel trajectories, environmental variables, and regulatory events, we examine spatiotemporal patterns of fishing effort relative to seasonal bans, extreme weather, and quota systems. Key findings reveal: (1) post-moratorium (September–December) concentration of fishing intensity, peaking in November; (2) significant weather-driven …

Face-To-Face Versus Distance Education In Competence-Based Maritime Education And Training, Clara Borén, Marcella Castells-Sanabra, Manel Grifoll

Journal of Marine Science and Technology–Taiwan

The COVID-19 pandemic created significant challenges for higher education, with universities being shut down and face-to-face teaching and assessment shifting to an online format. This presented an opportunity to focus on the continuity of learning through distance education, especially in the maritime context, due to the on-board training requirements, and even though five years have passed, this shift has reshaped the education landscape, proving that remote learning has come to stay.

At the Barcelona School of Nautical Studies, we have developed a weather routing software in the framework of distance education and teaching innovation. We conducted a teaching trial alternatively …

Gamma Belief Functions And Fuzzy Sets And Application To Combining Predictive Models, Liping Liu

University Research

Extending classic finite frameworks to continuous settings, this paper proposes the concept of gamma belief functions and gamma fuzzy sets. It shows that both the combination of gamma belief functions and the intersection of gamma fuzzy sets remain within the gamma family, enabling their application in combining gamma probability judgments in decision making and gamma regression models in ensemble learning. Using both simulated and real datasets, and under both constant and varying dispersion assumptions, experimental results show that the combined gamma regression models closely approximate the reference models learned from the full datasets, aligning with the objectives of bootstrapping. Notably, …

Re: Comment Letter For Butte Priority Soils Operable Unit (Bpsou), Draft Intermediate (60%) Remedial Design Report, Blacktail Creek Riparian Actions (Dated March 13, 2026), Emma Rott

Silver Bow Creek/Butte Area Superfund Site

No abstract provided.

Automatic Glossing In Under-Resourced Languages: Case Studies In Bribri And Cook Islands Māori, Carter D. Anderson

Linguistics Undergraduate Senior Theses

Interlinear glossing is a major task in Indigenous language documentation. In this paper, I explore how effectively two Large Language Models, ByT5 and Gemini 2.5 Flash, can produce interlinear glossed text. I also examine how prompting an LLM with different types of information (dictionary entries, other training samples, and translations) can augment model performance. I apply these models to two under-resourced Indigenous languages: Bribri, which is morphologically complex from Costa Rica, and Cook Islands Māori, which has a simpler morphology and is from the Cook Islands in the Pacific Ocean. ByT5 exhibits much better performance when glossing Cook Islands Māori …

Assessing The Origin Of The Talus Deposit In King Ravine In The Presidential Range Of New Hampshire, Margaret M. Keating

Earth Sciences Undergraduate Senior Theses

The timing and mechanisms driving the retreat of the Laurentide Ice Sheet (LIS) after the Last Glacial Maximum (LGM; 26-19 ka) in the Presidential Range of the White Mountains, New Hampshire, remains inconsistently interpreted. King Ravine, located within the Presidential Range, contains a landform that preserves geomorphic evidence relevant to this debate. Here, I investigate the origin of this landform by assessing its age, surficial morphology, and internal structure. I present a topographic analysis of King Ravine, including a detailed new geomorphic map of the drainage basin and deposit. To constrain the timing of landform formation and activity, I determined …

Exploring Paleoproterozoic Marine Redox States In A Young Banded Iron Formation From The Needle Mountains, Jaya Sharma

Earth Sciences Undergraduate Senior Theses

Banded iron formations (BIF) are redox-dependent chemical sedimentary rocks that have served as proxies for marine and atmospheric oxygen and iron conditions throughout Earth's history. While abundant in the Archean and early Proterozoic, the widespread deposition of BIFs declined following the Great Oxidation Event, as oxygenated surface waters prevented the accumulation of ferrous iron. However, the Needle Mountains, located in southwestern Colorado, house an anomalously young BIF deposited between 1786±10 Ma - 1801±6 Ma given existing geochronology. Here, we use rare earth element (REE) systematics, bulk chemistry, and petrographic analysis to determine the depositional conditions under which this BIF was …

Pitching Fwar Vs Bwar As Predictors Of Team Success In The Mlb Regular Season, Jason Lee

University Honors Theses

Pitching Wins Above Replacement (WAR) is an area of sabermetrics capable of being used to predict regular season success in Major League Baseball. Fangraphs WAR (fWAR) and Baseball Reference WAR (bWAR) were used to construct regression models to predict regular season winning percentage, to build logistic models to establish a relationship between pitching WAR and the probability to win an individual regular season game, and to overlay density plots to consider WAR accumulation by pitching role and observe the difference of impact between starter and relief pitchers. This research finds that while pitching fWAR and pitching bWAR are both statistically …

Assessing Spatial Distribution Of Plausible Earthquake-Triggered Landslides Relative To Ca. 1000 Ybp Earthquake On The Gales Creek Fault, Jacqueline Rheingans

University Honors Theses

Modern coseismic landslides cluster near the surface trace of the fault that ruptured, suggesting that clustering of paleolandslides may record information about paleoearthquakes. This study documents the pattern of prehistoric landslides near the Gales Creek Fault (GCF), northwest Oregon, to assess whether it reflects an independently known paleoearthquake. The GCF is located northwest of Portland through the Cascadia fore-arc basin where landslides are often triggered by intense rainfall during the wet season or potentially by seismic shaking on the megathrust or on crustal faults. In order to understand how a seismic event on the GCF will impact landslide distribution I …

Custom Sbc Gps Tracking And Geocaching Carputer Software Development And Implementation, Joshua A. Davis

University Honors Theses

This thesis argues that hardware-integrated capstone projects develop software engineering skills that traditional coursework cannot replicate. A team of eight developers built a GPS tracking system on a Raspberry Pi 4 over two academic terms, integrating real-time position streaming, the APRS amateur radio protocol for network-independent location sharing, and PostGIS spatial queries for "new road" detection. The system implements SmartBeaconing for adaptive GPS data reduction, achieving approximately 80% storage savings while preserving route fidelity. The project exposed challenges absent from classroom assignments: hardware debugging without stack traces, cross-layer integration failures, and coordination overhead when deploying to unfamiliar architecture; demonstrating that …

Synthetic Data Generation For Healthcare Machine Learning: A Case Study In Vital Signs And Diagnostic Predictions, Sellappan Palaniappan, Rajasvaran Logeswaran, Kasthuri Subaramaniam, Oras Baker, Bui Ngoc Dung

Journal of Informatics and Web Engineering

Healthcare Machine Learning (ML) applications face significant challenges in accessing high-quality training data due to stringent privacy regulations, institutional data silos, and concerns over patient confidentiality. This paper explores synthetic data generation as a viable and privacy-preserving alternative to real patient data for developing ML models in healthcare settings. We present techniques for generating realistic vital signs data including body temperature, blood pressure, heart rate, respiratory rate, and oxygen saturation according to appropriate statistical distributions. In addition, we demonstrate how synthetic datasets generated can be used to train diagnostic prediction models. The generated datasets were applied to multiple diagnostic prediction …

Ai-Powered File Security System With Facial Biometrics, Qr Code, And Otp Verification, Xinyi Loi, Ming Man Chan, Yuk Yii Chan

Journal of Informatics and Web Engineering

Commonly used file-sharing systems suffer from impermissible security risks such as impersonation, credential theft, and QR code exploitation, which can expose sensitive information. This paper describes and evaluates an AI-driven secure file-sharing system designed to alleviate these risks through a multi-layered authentication strategy. The complete system involves AES-128 encryption, unique QR code referencing, live facial recognition, and One-Time Password (OTP) confirmation. In our proposed workflow, the sender encrypts a reference to the file and links it to facial data of the recipient, if applicable, and embeds this information into a QR code, that is dynamically generated during the send operation. …

An Integrated Yolov8 Based System For Real Time Vehicle Tracking And Identification Using Cctv Footage, Riya Jain, Piyush Kumar, Ashwani Kumar Dubey, Angela Amphawan, Neo Tse Kian

Journal of Informatics and Web Engineering

Vehicle identification and tracking play a critical role in security and surveillance, supporting the efforts of law enforcing authorities to prevent crime, monitor traffic, and enforce the vehicle rules and regulations. This research presents a proof-of-concept system that employs YOLOv8 for real time vehicle detection and car type classification using live or recorded CCTV footage. YOLOv8 was chosen after rigorous comparison, it was chosen for its ease of use and efficient image training capabilities, making it well suited for rapid prototyping that the concept needed. The system is designed to extract key vehicle features, including colour, model of the vehicle, …