Physical Sciences and Mathematics Commons^™

Dgen: A Dynamic Generative Encryption Network For Adaptive And Secure Image Processing, Mohammed Rajih Jassim, Qusay M. Salih, Ghada Al-Kateb

Mesopotamian Journal of Computer Science

Cyber-attacks keep growing. Because of that, we need stronger ways to protect pictures. This paper talks about DGEN, a Dynamic Generative Encryption Network. It mixes Generative Adversarial Networks with a key system that can change with context. The method may potentially mean it can adjust itself when new threats appear, instead of a fixed lock like AES. It tries to block brute‑force, statistical tricks, or quantum attacks. The design adds randomness, uses learning, and makes keys that depend on each image. That should give very good security, some flexibility, and keep compute cost low. Tests still ran on several public …

Lightweight Deep Reinforcement Learning Model For Energy-Efficient Resource Allocation In Edge Computing, Ghassan A. Abed, Mohanad A. Al-Askari

Mesopotamian Journal of Computer Science

A lightweight digital twin model for a single 6G cell operating in the D-band (140 GHz) with a 1 GHz bandwidth is presented in this work with the goal of assessing the cell's capacity, coverage, and terminal time in order to support extended reality (XR) applications. With a tangent dispersion of 3 dB and a path exponent of n = 2.2, the model is based on the free-space loss equation as per ITU-R Recommendation P.525. The instantaneous capacity is determined using the Shannon-Hartley theorem. Three XR sessions are created every minute using a Poisson method, and their durations are determined …

Agentic Ai-Enhanced Virtual Reality For Adaptive Immersive Learning Environments, Indra Kishor, Udit Mamodiya, Mohammed Almaayah, Amer Alqutaish, Rami Shehab, Theyazn H. H. Aldhyani

Mesopotamian Journal of Computer Science

Immersive learning using Virtual Reality (VR) has gained prominence for delivering experiential, engaging education. However, most VR learning environments lack real-time adaptability, personalization, and cognitive responsiveness. This study presents an Agentic AI-enabled VR framework that autonomously adjusts pedagogical content, interaction style, and challenge level based on learner behavior, emotions, and performance feedback. The proposed system integrates a reinforcement learning-based agent with a virtual reality module to form an intelligent tutor capable of independent decision-making. A neuro-symbolic model processes multi-modal feedback (gesture, speech, gaze, performance) to determine context-aware pedagogical strategies. The system employs a self-evolving curriculum logic that adapts in real …

Foundation Models Boost Low-Level Perceptual Similarity Metrics, Abhijay Ghildyal, Nabajeet Barman, Saman Zadtootaghaj

Computer Science Faculty Publications and Presentations

For full-reference image quality assessment (FR-IQA) using deep-learning approaches, the perceptual similarity score between a distorted image and a reference image is typically computed as a distance measure between features extracted from a pretrained CNN or more recently, a Transformer network. Often, these intermediate features require further fine-tuning or processing with additional neural network layers to align the final similarity scores with human judgments. So far, most IQA models based on foundation models have primarily relied on the final layer or the embedding for the quality score estimation. In contrast, this work explores the potential of utilizing the intermediate features …

Osa-Diff: An Origin Sampling Based Adversarial Attack Using Diffusion Models, Shayan Jalalipour, Banafsheh Rekabdar

Computer Science Faculty Publications and Presentations

Diffusion models are becoming an increasingly popular emerging technology, however their use in adversarial attacks remains a scarcely explored topic. We show that diffusion models can be used to create end-to-end hidden adversarial perturbations with high rate of success, and propose a novel diffusion based adversarial attack that allows for substantially faster training time (through improved convergence on high quality images) and with substantially less computational overhead than typical diffusion model training

Polynomial-Time Constant-Approximation For Fair Sum-Of-Radii Clustering, Sina Nezhad, Sayan Bandyapadhyay, Tianzhi Chen

Computer Science Faculty Publications and Presentations

In a seminal work, Chierichetti et al. [20] introduced the (t,k)-fair clustering problem: Given a set of red points and a set of blue points in a metric space, a clustering is called fair if the number of red points in each cluster is at most t times and at least 1/t times the number of blue points in that cluster. The goal is to compute a fair clustering with at most k clusters that optimizes certain objective function. Considering this problem, they designed a polynomial-time O(1)- and O(t)-approximation for the k-center and the k-median objective, respectively. Recently, Carta et …

Retrieval-Augmented Feature Generation For Domain-Specific Classification, Xinhao Zhang, Jinghan Zhang, Fengran Mo, Dakshak Keerthi Chandra, Yu-Zhong Chen, Fei Xie, Kunpeng Liu

Computer Science Faculty Publications and Presentations

Feature generation can significantly enhance learning outcomes, particularly for tasks with limited data. An effective way to improve feature generation is to expand the current feature space using existing features and enriching the informational content. However, generating new, interpretable features usually requires domain-specific knowledge on top of the existing features. In this paper, we introduce a Retrieval-Augmented Feature Generation method, RAFG, to generate useful and explainable features specific to domain classification tasks. To increase the interpretability of the generated features, we conduct knowledge retrieval among the existing features in the domain to identify potential feature associations. These associations are expected …

“Lost-In-The-Later”: Framework For Quantifying Contextual Grounding In Large Language Models, Yufei Tao, Adam Hiatt, Rahul Seetharaman, Ameeta Agrawal

Computer Science Faculty Publications and Presentations

Large language models (LLMs) are capable of leveraging both contextual and parametric knowledge but how they prioritize and integrate these sources remains underexplored. We introduce CoPE, a novel framework for systematically quantifying contextual grounding in LLMs. CoPE distinguishes between contextual knowledge (CK) and parametric knowledge (PK), enabling fine-grained attribution across languages and tasks. Using our newly created MultiWikiAtomic dataset in English, Spanish, and Danish, we analyze how LLMs integrate context, prioritize information, and incorporate PK in open-ended question answering. We find that across models and languages, only around 50 to 76 percent of outputs are grounded in the given context, …

Scitopic: Enhancing Topic Discovery In Scientific Literature Through Advanced Llm, Pengjiang Li, Zaitian Wang, Xinhao Zhang, Ran Zhang, Lu Jiang, Pengfei Wang, Yuanchun Zhou

Computer Science Faculty Publications and Presentations

Topic discovery in scientific literature provides valuable insights for researchers to identify emerging trends and explore new avenues for investigation, facilitating easier scientific information retrieval. Many machine learning methods, particularly deep embedding techniques, have been applied to discover research topics. However, most existing topic discovery methods rely on word embedding to capture the semantics and lack a comprehensive understanding of scientific publications, struggling with complex, high-dimensional text relationships. Inspired by the exceptional comprehension of textual information by large language models (LLMs), we propose an advanced topic discovery method enhanced by LLMs to improve scientific topic identification, namely SciTopic. Specifically, we …

Binoculars To Bytes: Development And Field Validation Of An Ai-Driven System For Avian Monitoring, Christian J. Dupree

Graduate Student Theses, Dissertations, & Professional Papers

Autonomous camera-trap arrays coupled with artificial-intelligence (AI) vision can lift bird monitoring beyond the spatial, temporal, and labor limits of traditional field surveys. We present Binoculars to Bytes (B2B), an open-source pipeline that turns 180–360° time-lapse imagery into analysis-ready avian data. At Freezout Lake Wildlife Management Area (Montana, USA) the system ran four-hour morning deployments during spring and fall migrations (2023–2024). A YOLO-NAS detector, incrementally refined with a “Specialized Localized Iterative Model” workflow, quadrupled local accuracy and, after confidence-based species-binning, cut false-positive rates in half. Daily AI species lists were benchmarked against contemporaneous eBird citizen-science checklists and recovered ≥ …

Structural Optimization And Stability Enhancement Of Uba(2) And Ntl9 Using The Computational Tool: Emcast, Bahar Ghazi Esfahani

Graduate Student Theses, Dissertations, & Professional Papers

This study enhances protein stability prediction by refining EmCAST (Empirical C-Alpha Stability Tool), a computational tool used to identify stabilizing mutations in protein domains. We applied EmCAST to two model proteins: the α-helical ubiquitin-associated domain 2 (UBA(2)) and the α/β N-terminal domain of ribosomal protein L9 (NTL9). Protein stability is critical for proper function, and mutations can lead to misfolding and disease. To validate EmCAST’s predictions, guanidine hydrochloride-induced denaturation was monitored using circular dichroism (CD) spectroscopy for both wild-type and mutant variants. Results guided refinements to EmCAST’s algorithm, incorporating new experimental datasets to improve accuracy. Additionally, we examined the assumption …

Kinetic Grain Growth In Firn Induced By Meltwater Infiltration On The Greenland Ice Sheet, Kirsten Leigh Gehl

Graduate Student Theses, Dissertations, & Professional Papers

The microstructure of polar firn governs its porosity, permeability, and compaction rate, and is therefore important to studies of surface elevation change, heat and gas exchange, and meltwater infiltration on ice sheets. Previous work on high-elevation, dry-firn has identified two distinct atmospheric drivers of kinetic grain growth in deposited snow, but both mechanisms result in near-surface mm-scale layers. In this work, we show that meltwater infiltration through surface wetting fronts and deep preferential flow in the Greenland Ice Sheet (GrIS) percolation zone leads to the formation and preservation of centi- to decimeter scale layers of faceted firn, ranging from faceted …

Modeling Neighborhoods As Fuel For Wildfire, Bryce Alan Young

Graduate Student Theses, Dissertations, & Professional Papers

Wildfire models drive billions of dollars in risk mitigation efforts. However, the modeling community currently lacks a representative fuelscape on which to base simulations of fire spread in the built environment and the wildland-urban interface (WUI) where vegetation and structures act together as fuel for wildfire. This thesis advances wildfire risk modeling by addressing the underdeveloped representation of the built environment in existing frameworks. By identifying inconsistencies in how structure and defensible space features are defined and used across empirical studies, predictive indices, and fire spread models, this research lays the groundwork for standardized modeling approaches and feature selection (Chapter …

Bridging The Gap Between Plot-Level And Landscape-Scale Analysis For Wildfire Risk Assessment, Vanessa Leigh Niemczyk

Graduate Student Theses, Dissertations, & Professional Papers

Remote sensing technology has advanced greatly over the past couple of decades proving its ability to aid in wildfire risk assessment and improve our understanding of forest structure and fuel inventory across the landscape. While some aerial and satellite sensors perform better than others, they all have a common weakness, their reduced ability to capture understory fuels with high detail. Terrestrial laser scanning is an emerging solution due to its understory perspective. This research leverages the beneficial aspects of both terrestrial laser scanning and various aerial- or satellite-based remote sensing platforms (aerial laser scanning, digital aerial photogrammetry, and Sentinel-2) to …

Linking Empirical Data And Numerical Simulation To Characterize Dynamic Fire Behavior Associated With Interacting Firelines, Marta Sergeevna Jerebets

Graduate Student Theses, Dissertations, & Professional Papers

Understanding fuel pattern-fire process relationships is key for predicting fire behavior and effects with follow-on benefits to proactive fire management and model validation. To characterize dynamic fire behavior, this thesis leverages empirical data and numerical simulation through two complementary studies.

In the first study, longwave thermal sensors aboard unmanned aerial systems (UAS) were used to capture fine-scale fire behavior in two experimental grass burns. A novel paired design was used to quantify the effects of fuel arrangement on fire behavior with 3.66 m diameter treatments cut to a height of 0.15 m. The treatments ephemerally reduced fire rate of spread …

Investigating The Impact Of Aerial Firefighting On Rate Of Wildfire Spread, Lindsay Ann Wiard

Graduate Student Theses, Dissertations, & Professional Papers

Aerial retardant drops are widely used in wildfire suppression, yet their effectiveness in slowing fire spread remains difficult to quantify at scale. This study evaluates the impact of aerial suppression on wildfire rate of spread (ROS) using a modeling framework that incorporates both observed (real) and counterfactual (synthetic) drop locations from a sample of 62 wildfires in Oregon. Synthetic drops were generated to simulate a no-suppression baseline, allowing us to compare changes in ROS in the presence and absence of suppression. We trained two random forest classifiers: one using both real and synthetic drops (the full model), and another using …

Integrating Spatial Statistics And Decision Analysis For Wildfire Risk Mapping: A Case Study Of The Kootenai National Forest, Elijah Kordieh Mensah

Graduate Student Theses, Dissertations, & Professional Papers

Wildfire risk in the western United States has intensified in recent decades due to intersecting forces of climate change, systematic fire suppression, and expanding human settlement in fire-prone regions. This thesis presents a comprehensive spatial assessment of wildfire risk in the Kootenai National Forest (KNF) and its surrounding landscape in northwestern Montana, integrating wildfire likelihood, suppression difficulty, evacuation vulnerability, and building exposure into a unified spatial statistical framework. The research addresses a critical gap in spatial wildfire risk modeling by focusing on the need to assess fire threats in relation to local operational constraints and community vulnerabilities.
Drawing on geospatial …

Alive For The Wake, Anna Kate Powell

Graduate Student Theses, Dissertations, & Professional Papers

In the year 2046, bacteria strain B. bibitor has rendered all petroleum hydrocarbon compounds unrefinable. Paramedic Amanda Bailey finds hope in the resilience of humanity but struggles when she learns the powers that be would rather concentrate their efforts on fighting B. bibitor—restoring infected hydrocarbons to their previously refinable (and profitable) nature—rather than investing in already-developed clean firm energies and renewable material manufacture.

Fluvial Geomorphic Resilience In Alluvial Mountain Rivers To Large Floods, Riley T. Henson

Graduate Student Theses, Dissertations, & Professional Papers

As hydroclimatic extremes intensify, large floods increasingly reshape river systems and threaten infrastructure. Fluvial geomorphic resilience, defined as a river’s resistance to, and recovery from, flood disturbance, offers a useful framework for understanding river response and guiding management. We assessed resilience at two scales: (1) a detailed geomorphic and hydraulic analysis of the June 2022 flood on East Rosebud Creek, Montana, which was part of broader flooding in the Greater Yellowstone Ecosystem, and (2) a broad-scale comparison of 12 flood-impacted alluvial mountain rivers across North America and Europe. We quantified resistance using active channel width ratios (W_R) and …

Investigating Non-Invasive Survey Methods And Building A Predictive Occupancy Model For Harlequin Ducks (Histrionicus Histrionicus) In Western Montana And Northern Idaho, Holli Ann Holmes

Graduate Student Theses, Dissertations, & Professional Papers

Harlequin Ducks (Histrionicus histrionicus; HADU) are an imperiled sea duck that breeds along white-water mountain streams, but challenges in surveying them have made it difficult to understand population trends in their Northern Rocky Mountain breeding population. In this study we had two objectives: (1) evaluate non-invasive survey methods for detecting HADU on their breeding grounds in Western Montana and Northern Idaho, and (2) develop an occupancy model to predict potential breeding stream use for HADU across the region. First, we assessed the efficacy of ground-based foot surveys (GBS), environmental DNA (eDNA), and camera traps to detect breeding HADU during …

Changes In Efficiency And Magnitude Of Recharge Due To Climate Change For Headwater Catchments In The Mountain West, Harrison Mcgillen

Graduate Student Theses, Dissertations, & Professional Papers

Changes to groundwater recharge and stream flow were simulated from downscaled CMIP6 global climate models in a headwater catchment of the Blackfoot River, MT. A calibrated, 3-D, coupled surface and groundwater GSFLOW model, forced with minimum and maximum temperature, and precipitation from a downscaled CMIP6 simulation was used to quantify groundwater recharge and stream flow. The model was calibrated against measured groundwater head and stream flow using observed temperature and precipitation boundary conditions. The calibrated model was then forced with downscaled temperature and precipitation from the CANESM5 model for the SSP245, 370 and 585 scenarios, as well as a manual …

Depth Distribution Of Liquid Water In Greenlands Firn Layer Based On L-Band Radiometry, A Snow Physics Model, And Machine Learning, Taylor D. Moon

Graduate Student Theses, Dissertations, & Professional Papers

The accurate estimation of the amount of meltwater retained by Greenland’s firn layer and its distribution over depth is essential for assessing both the current and future mass balance of the Greenland Ice Sheet. When meltwater infiltrates the firn, it exists as a quantifiable volume of liquid water, measurable as the one-dimensional liquid water amount, hereafter referred to as LWA. A promising new tool for estimating LWA is L-band passive microwave satellite radiometry, which can provide twice-daily estimates using empirical models of firn microwave emission. However, this method faces two primary limitations. First, retrievals are untested against field observations and …

Mapping Mesic Resilience In Western Rangelands: Productivity Trends, Predictors, And Targeting Frameworks, Kristopher Robert Mueller

Graduate Student Theses, Dissertations, & Professional Papers

Mesic ecosystems in riparian areas and wetlands make up only 2% of western U.S. rangelands but support 80% of wildlife during dry periods, providing vital late-season forage and refuge when surrounding uplands dry out. Yet, mesic habitats are often overlooked in landscape-scale conservation planning. This thesis addresses two key gaps: understanding how mesic productivity responds to environmental change and developing frameworks to incorporate these habitats into current conservation strategies. Using 40 years of Landsat imagery (1984–2024), we examined long-term trends in late-season mesic vegetation productivity and its relationship with drought across the sagebrush biome. From 1984 to 2004, mesic productivity …

Impacts Of Fire-Driven Vegetation Changes On Evapotranspiration In North American Boreal Forests, Tyler Glen Albrethsen

Graduate Student Theses, Dissertations, & Professional Papers

A growing body of evidence suggests that land cover change could be impacting evapotranspiration (ET) trends more than the direct impacts of climate change (Dashti et al., 2022; Zhang et al., 2015). In boreal forests, one of the Earth’s most important carbon sinks (Kang et al., 2006), climate and disturbance-driven land cover change is progressing rapidly, threatening the resilience of conifer-dominated forests (Baltzer et al., 2021; Massey et al., 2023). The spread of deciduous trees and the changing structure of boreal forests likely have large implications for the magnitude and seasonality of ecosystem water fluxes (Young-Robertson et al., 2016). This …

Ecology-Driven Machine Learning For Predicting Risk Of Spread For Aquatic Invasive Species., Leif Eric Howard

Graduate Student Theses, Dissertations, & Professional Papers

Effective management of aquatic invasive species (AIS) requires predictive tools that leverage environmental and ecological data to forecast spread into uninvaded areas. However, species distribution models (SDMs) often struggle to capture key invasion processes such as dispersal, propagule pressure, and biotic interactions. This thesis integrates ecological, dispersal, and environmental predictors within a machine learning framework to enhance AIS risk prediction and model transferability across space.

In Chapter 2, 61 novel predictors were developed for 2,203 Minnesota lakes to model zebra mussel (Dreissena polymorpha) occurrence, incorporating variables related to propagule pressure (e.g., boat visitation), community composition (e.g., fish species …

Campesino World-Building: Stories Of Migration, Memory, And Care In The Tension Of The Colonial Frame, Chelsea Ríos Gómez

Cal Poly Humboldt theses and projects

Campesinos, often reduced to the label of “farmworkers”, are too often seen only through the lens of their labor within profit-driven systems. This thesis disrupts that limited narrative by illuminating the rich, expansive nature of vivencias campesinas, ways of being that transcend capitalist measures of value. These vivencias are lived, embodied experiences rooted in campesino lifeways, encompassing cultural, spiritual, and relational practices that nurture and sustain life.

My research is guided by four central questions: (1) In what ways have campesinos in pixley resisted exploitative labor systems, and how can this resistance be understood as a form of care labor? …

Exploring Coastal Spaces Of Tribal Sovereignty And Self-Determination In Ancestral Waters Proposed For Offshore Wind Development, Alyssa Suarez

Cal Poly Humboldt theses and projects

Proposed offshore wind development off California’s North Coast raises novel questions about Tribal sovereignty offshore. This project specifically engages claims, assuming the Yurok Tribe’s framing that offshore wind development poses a coastal hazard. The research followed a co-production framework. The research questions were co-designed with the Yurok Tribe and the research itself governed by a Data Sharing Agreement. The work presents findings from 13 semi-structured interviews with individuals and experts navigating sovereignty in the offshore wind space, including government representatives; lawyers; legal researchers; academics; foundation staff; journalists; and Humboldt Bay community members. Participants identified several key tensions surrounding claims of …

North American River Otter (Lontra Canadensis) Prey Selection Of Wintering Waterbirds In A Wetland Complex, Gina Yael Culver

Cal Poly Humboldt theses and projects

Generalist predators are assumed to take a wide variety of prey throughout an annual cycle, often switching between food sources depending on availability. Prey selection is often studied by way of variables such as prey age class, body size, and various topics related to anti-predator behaviors, all variables used to measure how predators target prey. Following prior studies in coastal Northern California documenting that river otters (Lontra canadensis) had selected waterbirds as prey, we identified undigested prey items in scat to evaluate prey selection of waterbirds at the Arcata Marsh and Wildlife Sanctuary (AMWS). We compared feathers found …

Turbo Fladry Fencing To Mitigate Short-Term Agricultural Damage Caused By Wild Pigs, Michael P. Glow, Jesse A. Tussing, Johnathan G. Walls, Robert B. Hyle, Joseph M. Halseth, Nathan P. Snow, Kurt C. Vercauteren

Human–Wildlife Interactions

Wild pigs (Sus scrofa) are a destructive invasive species that damage agricultural crops in many regions of the world. Lethal techniques to reduce populations are most often used to mitigate damage, but additional techniques are also needed in conjunction with lethal tools to protect resources in some settings (e.g., where lethal control is not practical or possible to achieve population elimination). While additional tools such as deterrents or temporary fencing may provide short-term relief, few effective options are currently available because habituation by wild pigs often reduces long-term effectiveness. Turbo fladry, a type of electric fence, is an …

When Emotions Flare: Solar Rhythms, Emotions And Cycles Of Political Revolution, Andreas Hernandez, Carolina Zilli Vieira, Alexandra Smith, Rebecca Olson

Geography and Environmental Studies Faculty Publications

Revolutions are among the most transformative events in human history. Analyzing 395 revolutionary episodes from 1900 to 2014, we find that major waves cluster around solar maxima - periods of intensified geomagnetic activity. Drawing on biomedical research linking geomagnetic disturbances to cardiovascular and stress regulation, and thus to emotional states, we propose that solar cycles modulate the affective ecologies within which revolutions arise. Solar activity may accelerate, shape, and amplify revolutionary cycles by heightening emotional climates and tipping fragile systems toward mass mobilization. This reframes revolutions as planetary events - entanglements of political conflict, embodied emotion, and cosmic …