Digital Commons Network^™

Machine Learning For Asphaltene Polarizability: Evaluating Molecular Descriptors, Arun K. Sharma, Owen Mcmillan, Selsela Arsala, Supreet Gandhok, Rylend Young

Biology, Agriculture and Chemistry Faculty Publications and Presentations

Asphaltenes are complex polycyclic organic molecules in crude oil that readily aggregate and precipitate under varying thermodynamic conditions. Their structural heterogeneity influences key physicochemical properties, including solubility, stability, and reactivity. Molecular polarizability, a crucial property governing intermolecular interactions and electronic behavior, remains challenging to predict due to this structural diversity. This study employs machine learning models to predict isotropic polarizability using two sets of molecular descriptors: WHIM and GETAWAY. A dataset of 255 asphaltene structures was analyzed using stratified sampling, generating 10 independent training (80 %) and testing (20 %) splits. The Wolfram Language’s Predict function evaluated multiple machine learning …

Aggregator Zone Selection For Ev Smart Controls Based-On Ml Clustering Of Grid Strength, Distance, And Charging Homogeneity, Rosemary E. Alden, Sam H. Lowe Ii, Dan M. Ionel

Electrical and Computer Engineering Graduate Research

Smart electric vehicle (EV) charging control methods from a central utility hub often require communication infrastructure over a large service area of electric power distribution systems with a large number of nodes. Industry standards such as Open Charge Point Protocol (OCPP) 2.1 have evolved to include topologies for local controllers to the individual chargers, i.e. EV aggregator zones. A machine learning (ML) application of k-means clustering is proposed to establish zones for coordination of EV charging based on grid strength and EV owner decision-making to charge per day. Very large-scale distribution networks including the IEEE 123 and 8500 benchmark feeders …

Hybrid Fea And Meta-Modeling For De Optimization Of A Pm Stator-Excited Motor With A Reluctance Rotor, Oluwaseun A. Badewa, Dan M. Ionel

Electrical and Computer Engineering Graduate Research

This paper presents an innovative method for designing high-performance electric motors by integrating machine learning (ML) based meta-modeling with a differential evolution (DE) optimization algorithm. The approach utilizes finite element analysis (FEA) data to train the ML meta-model, allowing for efficient optimization of high-power-density machines, such as the reluctance rotor and permanent magnet (PM) stator combined excitation motor, which is characterized by nonlinearities. The meta-modeling process employs an Artificial Neural Network (ANN) with 3 hidden layers and uses the motor’s geometrical variables as inputs. The accuracy of the meta-model in capturing the nonlinear relationships between design parameters, core losses, and …

Modeling Depressive Patterns In Italian Discourse: Insights From Natural Language Processing, Annapia Borraccino

Dissertations, Theses, and Capstone Projects

Depression does not always speak in obvious ways. It often hides in silence, in brief pauses, and in the quiet texture of language. While the prosodic and linguistic characteristics of depression have been extensively studied in English and other languages, they remain underexplored in Italian, with this being a first focused analysis. This study investigates how depression manifests in the speech of Italian speakers, using transcriptions from a clinical corpus. Drawing on emotional, structural, and cognitive-linguistic features, including first person verb use, sentence complexity, verb tense, and negation, the study evaluates how these markers differ between individuals with and without …

Irnet: Immunotherapy Response Prediction Using Pathway Knowledge-Informed Graph Neural Network, Yuexu Jiang, Manish Sridhar Immadia, Duolin Wang, Shuai Zeng, Yen On Chan, Jing Zhou, Dong Xu, Trupti Joshi

Biomedical Sciences

Introduction: Immune checkpoint inhibitors (ICIs) are potent and precise therapies for various cancer types, significantly improving survival rates in patients who respond positively to them. However, only a minority of patients benefit from ICI treatments.

Objectives: Identifying ICI responders before treatment could greatly conserve medical resources, minimize potential drug side effects, and expedite the search for alternative therapies. Our goal is to introduce a novel deep-learning method to predict ICI treatment responses in cancer patients.

Methods: The proposed deep-learning framework leverages graph neural network and biological pathway knowledge. We trained and tested our method using ICI-treated patients’ data from several …

Developing Machine Learning Models And Graphene-Based Flexible Humidity And Temperature Sensors For Machine-Learning-Assisted Sensing, Seth Hajian

Dissertations

Flexible sensor technology has recently gained tremendous momentum in both academic research and industrial applications, transitioning from conceptual frameworks to practical implementations across diverse fields. This remarkable advancement can be attributed to several converging factors, including the maturation of nanomaterial science, the advancements of machine learning algorithms, and the critical demand for intelligent sensing solutions in healthcare, environmental monitoring, and industrial automation. The growing emphasis on personalized medicine and real-time health monitoring, accelerated by global health challenges, has further highlighted the necessity for accurate, cost-effective, and adaptable sensing platforms. This dissertation presents the fulfillment of three interconnected research projects focused …

Development Of A Machine-Learning-Driven Digital Teaching Assistant That Utilises Student Engagement Data To Improve Access To And Success In K-12 Stem Education, Kelly Shreeve, Kathryn Jessen Eller, Michael Cassidy, Beth Price, Brandy Jackson, Anthony M. Perry, Leo Celi, Ismini Lourentzou, Luk Hendrik

Teaching, Leadership & Professional Practice Faculty Publications

Student engagement is a key predictor of academic achievement and is closely linked to career awareness, interest, and preparedness. Measuring student engagement during STEM learning is challenging for teachers, given the dynamic and ever-changing nature of these learning environments. Even when engagement data can be collected, leveraging this information to refine and personalise instruction requires significant experience and time. To address this, we are developing Scoutlier EngagEd, a digital teaching assistant that embeds in existing Learning Management Systems (LMS) to automatically and invisibly gather multidimensional data on student engagement and performance during STEM learning. These data are being leveraged to …

Towards Explainable Ai On Graph Neural Networks: Xaig, Jiaxing Zhang

Dissertations

In the evolving landscape of artificial intelligence (AI), Graph Neural Networks (GNNs) have garnered growing prominence for their adeptness in processing graph-structured data. Despite this, the interpretability of their predictions often remains elusive. The demand for transparency and explainability in complex prediction models has reached unprecedented levels. To address this, post-hoc instance-level explanation techniques have emerged, aiming to unveil the rationale behind GNN predictions. These techniques endeavor to unearth substructures that elucidate the predictive behavior of trained GNNs.

This dissertation embarks on an exploration of Explainable AI (XAI) technologies within the realm of GNNs. Amid the challenges posed by the …

Machine Learning For Preeclampsia Prediction: Enhancing Screening In Primary Health Care, Dwirani Amelia, Asri Adisasmita, Kemal N. Siregar, Detty Siti Nurdiati

Kesmas

Preeclampsia is a leading cause of maternal morbidity and mortality worldwide, with early detection being critical for reducing adverse outcomes. This study aimed to develop a machine learning model for predicting the risk of preeclampsia using readily available maternal characteristics such as body mass index, mean arterial pressure, and clinical history of hypertension or diabetes mellitus. Secondary data from 2,250 pregnancies were analyzed, addressing challenges such as missing data and class imbalance through preprocessing. Various algorithms, including support vector machines, random forest, and logistic regression, were evaluated. Herein, a support vector machines model with threshold adjustment showed the best performance, …

Benchmarking Machine Learning Models In Lesion-Symptom Mapping For Predicting Language Outcomes In Stroke Survivors, Deepa Tilwani, Christian O'Reilly, Nicholas Riccardi, Valerie L. Shalin, Dirk B. Den Ouden, Julius Fridriksson, Svetlana V. Shinkareva, Amit P. Sheth, Rutvik H. Desai

Communication Sciences and Disorders Faculty Articles and Research

Several decades of research have investigated the neural connections between stroke-induced brain damage and language difficulties. Typically, lesion-symptom mapping (LSM) studies that address this connection have relied on mass univariate statistics, which do not account for multidimensional relationships between variables. Machine learning (ML) techniques, which can capture these intricate connections, offer a promising complement to LSM methods. To test this promise, we benchmarked ML models on structural and functional MRI to predict aphasia severity (N = 238) and naming impairment (N = 191) for a cohort of chronic-stage stroke survivors. We used nested cross-validation to examine performance along …

Optimizing Concrete Strength: How Nanomaterials And Ai Redefine Mix Design, Dan Huang, Guangshuai Han, Ziyang Tang

Physics and Engineering Science

Nanomaterials and supplementary cementitious materials (SCMs) are typically used together in efforts to enhance the performance of concrete and mitigate the environmental impact of concrete construction. However, the complex interactions between nanomaterials, SCMs, and cement make concrete mix design a challenging, iterative, and labor-intensive process, often relying on trial-and-error experimentation. Machine learning (ML) offers an opportunity to better understand the influence of input parameters and to accelerate the optimization of mix designs through data-driven insights. This study proposes an open-source and easy-to-access framework, Canopy, to support the concrete research community in optimizing mix design. Using a dataset collected from the …

Clinician Attitudes And Perceptions Of Point-Of-Care Information Resources And Their Integration Into Electronic Health Records: Qualitative Interview Study, Marlika Marceau, Sevan Dulgarian, Jacob Cambre, Pamela M Garabedian, Mary G Amato, Diane L Seger, Lynn A Volk, Gretchen Purcell Jackson, David W Bates, Ronen Rozenblum, Ania Syrowatka

Faculty, Staff and Students Publications

Background: Electronic health records (EHRs) are widely used in health care systems across the United States to help clinicians access patient medical histories in one central location. As medical knowledge expands, clinicians are increasingly using evidence-based point-of-care information (POCI) resources to facilitate clinical decision-making in medical practices. While these tools can improve clinical outcomes, few studies have assessed clinicians' opinions on integrating them with EHRs.

Objective: This study aims to assess clinicians' attitudes and the perceived value of POCI resources for finding medication- and disease-related information in clinical practice and their integration with EHRs.

Methods: Semistructured interviews were conducted with …

Course Portfolio: Csce 478/878 Introduction To Machine Learning, Hongzhi Guo

UNL Faculty Course Portfolios

This course portfolio was developed for CSCE 478/878: Introduction to Machine Learning, offered by the School of Computing at the University of Nebraska–Lincoln, as part of the Faculty-led Inquiry into Reflective and Scholarly Teaching (FIRST) project. The course provides an introduction to foundational machine learning concepts, models, and programming tools. It enrolls both undergraduate and graduate students from a wide range of academic backgrounds, including computer science, computer engineering, software engineering, biological systems engineering, chemical engineering, business-related fields, etc. Offered in Spring 2025, the course was peer-reviewed to support continuous improvement in teaching. The primary goals of this portfolio are …

Retracted: Robust Security System: A Novel Facial Recognition Optimization Using Coronavirus-Inspired Algorithm And Machine Learning, Saif Mohanad Kadhim, Johnny Koh Siaw Paw, Yaw Chong Tak, Shahad Thamear Abd Al-Latief

Iraqi Journal for Computer Science and Mathematics

Facial recognition has become an invaluable and rapidly advancing technology that plays a crucial role in various daily applications. From identity authentication to video surveillance, mobile payment, and even law enforcement and security measures. Despite the remarkable progress, facial recognition is still a dynamic research field and confronts several challenges. One of the main challenges is the high variability in facial images due to factors like facial expressions, lighting conditions, aging, and the presence of accessories. Additionally, the computational complexity and the time concerns surrounding face recognition systems have raised considerations that need to be addressed. This research presents a …

San Jose Urban Forest – An Open-Source Tree Canopy Surveying And Assessment Tool, Martin Alvarez Lopez, Manan Choksi, Sai Yaaminie Ganda, Vaibhavi Hiteshkumar Savani, Bernardo Flores

Faculty Research, Scholarly, and Creative Activity

The accurate monitoring of tree canopy coverage is crucial for social and environmental studies as well as city planning. Our goal is to provide a solution that utilizes available resources like yearly aerial imagery to generate tree data to track all tree canopy within a city such as San Jose, CA. By utilizing open-source software and publicly available data, we streamlined the process by eliminating the need for expensive equipment and time-consuming manual efforts associated with traditional methods. The methodology involves the conversion of aerial imagery datasets into geo-referenced TIFF format through a resampling process. To optimize efficiency, we strategically …

Vaccine Hesitancy In Türkiye: A Natural Language Processing Study On Social Media, Semi̇h Sari, Ulya Bayram

Turkish Journal of Electrical Engineering and Computer Sciences

Vaccine hesitancy is a significant public healthcare problem that is threatening everyone worldwide. Vaccine hesitancy has become more ingrained in Turkish society, mainly through social media. Unfortunately, reflections of this hesitancy are preventable deaths or permanent disabilities. Because of the uncontrolled spread of misinformation and disinformation on social media, Türkiye is facing a future health crisis. As a step towards preventing this crisis, our main objective is to use the power of artificial intelligence techniques on Turkish social media posts to detect antivaccine posts. Through this study, it will be possible to raise awareness about the importance of vaccines in …

Predicting Patient No-Show Rate In Healthcare To Improve Operational Excellence, Saeed Ahmed Al Wari

Theses

The challenge of integrating no-show predictive models into healthcare environments is complex and multifaceted, extending far beyond the distinct technical challenge of building accurate models. This thesis will discuss more on the subject of incorporating machine learning into healthcare. While algorithmic accuracy and model validation are critical, they are a mere sub-system of a much larger operational, technological, ethical, legal and human ecosystem. Predictive models face mass deployment in healthcare systems which are complex networks of stakeholders across strict regulatory frameworks and deeply embedded workflows (AlMuhaideb et al., 2019). However, for these models to result in long term and real …

Predicting Customer Churn In E-Commerce, Hind Tawfiq Kazim

Theses

Customer retention has become a critical focus for businesses seeking to sustain growth and profitability in an increasingly competitive market. In this thesis, advanced machine learning techniques are used to develop a data-driven churn prediction model for Majid Al Futtaim's customer base. In an e-commerce environment, the research focuses on identifying the key transactional and behavioral factors that influence customer churn, based on customer relationship management theory. The study is contextualized in the context of increased digital consumer engagement and high customer acquisition costs, which make retaining existing customers more cost- effective than acquiring new ones. Our primary research questions …

Hybrid Sentiment Analysis Of Police Social Media: Managing Big Data Impact On Youth, Alunood Yousuf Alamiri

Theses

The pervasive influence of social media has reshaped public communication, particularly for institutions like law enforcement agencies that rely on these platforms to engage with the community. As police departments increasingly utilize social media to disseminate information, interact with the public, and manage public relations, the need for effective sentiment analysis becomes crucial. Understanding how police messages are received by the public, especially by the youth, who are among the most active and impressionable users of social media, is essential for maintaining public trust and ensuring effective communication. This thesis, seeks to develop a comprehensive solution for this challenge. The …

Application Of Pu Learning In Detection Of Ddos Attacks, Gagana Sathya Narayana Prasad

Theses and Dissertations

The gcore radar 2024 says, the number of DDoS attacks has been increased by 46% in 12 months. Supervised and unsupervised techniques struggle detecting DDoS attacks due to the scarcity of labeled attack samples and an overwhelming presence of benign traffic. In contrast PU- Learning offers a promising solutions by dividing the data into positive and unlabeled data. This study explores the effectiveness of PU-learning in detecting DDoS attacks by comparing it with unsupervised methods. This method employs PU Bagging, Two Step method and auto-encoder based models to extract meaningful patters from network traffic data, utilizing CICDDoS2017 dataset for evaluation. …

Mfgat: Map-Free Trajectory Prediction With Graph Attention Networks For Autonomous Vehicles, Zehra Gunindi

UNLV Theses, Dissertations, Professional Papers, and Capstones

Accurate trajectory prediction is a key component for ensuring safe and efficient navigation of autonomous vehicles in complex traffic scenarios. While traditional methods rely heavily on high-definition (HD) maps, these approaches face significant challenges, including high costs, limited availability, and susceptibility to rapid obsolescence. This thesis proposes an end-to-end, map-free trajectory prediction model that leverages Graph Attention Networks (GAT) to dynamically capture spatial-temporal interactions among road agents, eliminating the need for HD maps.The research introduces UNLVTraj, a novel LiDAR-based dataset collected around the University of Nevada, Las Vegas campus, specifically along Cottage Grove Street, Harmon Avenue, and Maryland Parkway. This …

Efficient Eeg Epilepsy Classification And Feature Selections Based On Hellinger Distance, Muhammed Sadiq

Theses and Dissertations

Accurate and efficient detection of epileptic seizures from EEG signals remains a critical challenge due to high-dimensional data, class imbalance, and the limitations of standard classifiers. This thesis introduces two novel models to address these challenges. The first model presents a new classifier based on the Hellinger Distance, specifically designed to enhance discriminative capability and robustness against imbalanced datasets. By integrating the Hellinger Distance Classifier with Particle Swarm Optimization (PSO) for feature selection, this model significantly improves classification performance while reducing computational complexity. Experimental evaluations on the Bonn dataset demonstrate an accuracy of 96.25%, an F1-score of 97.74%, a recall …

Computational Complexity Of Soundness Verification For Neural Networks, Scott Sirri

McKelvey School of Engineering Graduate Student Theses & Dissertations

Neural networks are an increasingly ubiquitous tool in systems of varying complexity across a range of domains. While these tools can be used to learn and predict complex functions, their opaque nature limits the scope of their acceptable applications. In particular, a lack of performance guarantees means that they are unsuitable for safety-critical applications such as self-driving cars and scheduling systems. Neural networks trained to solve NP-complete problems, in particular, are unlikely to be able to solve the problem exactly. However, a weaker soundness guarantee may be sufficient for some systems, e.g., that positive instances of the problem may be …

Identification Of Subtypes Of Post-Stroke And Neurotypical Gait Behaviors Using Neural Network Analysis Of Gait Cycle Kinematics, Andrian Kuch, Nicolas Schweighofer, James M. Finley, Alison Mckenzie, Yuxin Wen, Natalia Sánchez

Physical Therapy Faculty Articles and Research

Gait impairment post-stroke is highly heterogeneous. Prior studies classified heterogeneous gait patterns into subgroups using peak kinematics, kinetics, or spatiotemporal variables. A limitation of this approach is the need to select discrete features in the gait cycle. Using continuous gait cycle data, we accounted for differences in magnitude and timing of kinematics. Here, we propose a machine-learning pipeline combining supervised and unsupervised learning. We first trained a Convolutional Neural Network and a Temporal Convolutional Network to extract features that distinguish impaired from neurotypical gait. Then, we used unsupervised time-series k-means and Gaussian Mixture Models to identify gait clusters. We tested …

Growing Through Simulated Realities: Proxy-Tuning Llms With Embodied Multi-Task Planning, Sirui Wang

Dartmouth College Master’s Theses

This thesis presents a method for teaching large language models (LLMs) to perform embodied multi-task planning using simulation data and a lightweight training technique called proxy-tuning. Traditional LLMs are strong in language understanding but struggle with tasks that require reasoning and planning in real-world environments. To address this, we construct a dataset based on VirtualHome, a simulator that models household activities using input commands. We extract individual tasks from the simulator and combine them into multi-task plans, optimizing for coherence and efficiency. These plans are formatted as Q&A pairs to train models to generate step-by-step instructions that complete multiple tasks …

Satellite Reorientation Using Reinforcement Learning Under Unknown Attitude Failure, Matthew Willoughby

Doctoral Dissertations and Master's Theses

This study presents a reinforcement learning (RL) approach for reestablishing communication with deep-space satellites under unknown attitude determination and control system (ADCS) failures. When traditional fault-tolerant control methods cannot restore signal, the proposed RL controller acts as a last-resort measure by autonomously reorienting the satellite’s antenna toward Earth while charging the battery via solar panels. A generic reward function, designed for the RL-based method, enables the controller to adapt to diverse failure scenarios, including severe actuator noise, misalignment, and complete actuator failure. Simulations are conducted in the Basilisk environment and trained with the tonic framework and demonstrate ranging capabilities of …

Improving The Reproducibility Of Deep Learning Software: An Initial Investigation Through A Case Study Analysis, Nikita Ravi, Abhinav Goel, James C. Davis, George K. Thiruvathukal

Computer Science: Faculty Publications and Other Works

The field of deep learning has witnessed significant breakthroughs, spanning various applications, and fundamentally transforming current software capabilities. However, alongside these advancements, there have been increasing concerns about reproducing the results of these deep learning methods. This is significant because reproducibility is the foundation of reliability and validity in software development, particularly in the rapidly evolving domain of deep learning. The difficulty of reproducibility may arise due to several reasons, including having differences from the original execution environment, incompatible software libraries, proprietary data and source code, lack of transparency, and the stochastic nature in some software. A study conducted by …

Estimating Pedestrian Crossing Times At Scramble Crossings Via Machine Learning And Agent-Based Modeling, Sho Takami

Honors Capstones

Scramble crosswalks differ from conventional crosswalks in their ability for pedestrians to cross diagonally. This research compares the average crossing times and investigates the walking behaviors that pedestrians adopt to produce the speediest times in the two crosswalk configurations. Identification of the most efficient set of walking behaviors is done through an agent-based model, whereas producing polynomials relating crossing times to the most prominent walking behaviors is done through regression algorithms in machine learning. With the combination of these two approaches, it is revealed that pedestrians must adopt a relaxed walking style to make each crosswalk configuration efficient. Additionally, between …

Data-Driven Dynamic Decision-Making Using Discrete Optimization And Supervised Machine Learning, Navid Rashedi

Dartmouth College Ph.D Dissertations

In recent years, the operations research community has developed data-driven optimization techniques to solve complex combinatorial problems with the aid of machine learning. This thesis contributes to these efforts by combining machine learning with optimization to expedite online decision-making, with applications in transportation and healthcare.

In the domain of airline operations recovery, the focus is on the aircraft recovery process—repairing disrupted schedules by minimizing overall disruption costs. Traditional exact methods are too time-consuming, while heuristic approaches often yield poor solution quality and lack generalizability across varying formulations. To address these challenges, this research employs supervised machine learning to identify near-optimal …

Evaluating And Applying Coherence Models: Llms As Coherence Evaluators And Neural Entity Grids As Noun Entity Shift Indicators, Elyas A. Amin

Theses and Dissertations

This thesis investigates computational approaches to discourse coherence through two interconnected studies. First, we evaluate the capabilities of Large Language Models (LLMs) as coherence evaluation models across real-world textual data domains. Our experiments with GPT-4o, GPT-3.5, Llama, and Mistral models reveal that while LLMs significantly outperform traditional entity grid methods (55% improvement in F1 scores), their performance varies substantially across domains and coherence levels. Chain-of-thought prompting unexpectedly decreased performance, and fine-tuning experiments demonstrated improved detection of low and high coherence texts but diminished performance on medium coherence cases. In our second study, we explore the practical application of entity grid-based …