Statistics and Probability Commons^™

Ms-Yolo: Infrared Object Detection For Edge Deployment Via Mobilenetv4 And Slideloss, Jiali Zhang, Thomas S. White, Haoliang Zhang, Wenqing Hu, Donald C. Wunsch, Jian Liu

Mathematics and Statistics Faculty Research & Creative Works

Infrared imaging has emerged as a robust solution for urban object detection under low-light and adverse weather conditions, offering significant advantages over traditional visible-light cameras. However, challenges such as class imbalance, thermal noise, and computational constraints can significantly hinder model performance in practical settings. To address these issues, we evaluate multiple YOLO variants on the FLIR ADAS V2 dataset, ultimately selecting YOLOv8 as our baseline due to its balanced accuracy and efficiency. Building on this foundation, we present MS-YOLO (MobileNetv4 and SlideLoss based on YOLO), which replaces YOLOv8's CSPDarknet backbone with the more efficient MobileNetV4, reducing computational overhead by 1.5% …

Optimizing Decision-Making In A Cerebral Palsy Model Using Reinforcement Learning, Richard Ampah

Pitzer Senior Theses

This study presents an original interdisciplinary investigation into how reinforcement learning (RL) can model motor and cognitive defects and potentially improve motor and cognitive functions in individuals with cerebral palsy (CP), a non-progressive neurological disorder that impairs movement and adaptability. Integrating computational neuroscience and machine learning, the research applies policy gradient methods and Markov Decision Processes (MDPs) to simulate adaptive learning in agents with and without CP-related constraints.

The central aim is to compare the cumulative rewards of optimal policies, derived from value iteration, and human-like learning policies using the REINFORCE algorithm, both with and without the Bellman baseline. The …

Local Limit Theorems On Finitely Generated Abelian Groups, Yutong Yan

Honors Theses

In this thesis, we classify the pointwise behavior of finite-range random walks on finitely generated abelian groups in terms of local limit theorems. Random walks are central objects of research in probability theory, and the theory has found applications in statistics, physics, and even card shuffling. One significant topic in this line of study is random walks on finitely generated groups. Starting from the pioneering work of G. Pólya and H. Kesten, random walks on finitely generated groups have been studied extensively. However, many notable results on the subject (local limit theorems, for example) make assumptions about periodicity and irreducibility …

Fault Tree Analysis For Robust Design, Jonathan Degroff, Gene Jean-Win Hou

Mechanical & Aerospace Engineering Faculty Publications

The objective of this research is to incorporate system failure into a robust design formation and solution process. The system failure referred to here will be built using fault tree analysis (FTA), which will take all lower-level failure events into consideration. Two examples are investigated here. One will directly treat the probabilities of the basis events as design variables, The other will be formulated in five different models: deterministic design optimization, the reliability index-based, the “and” gate-based, the “or” gate-based and the “inhibit” gate-based robust design. Their corresponding optimization solutions will be compared with each other. The post-optimality analysis of …

Comparative Study Of Single Imputation Techniques For The Prediction Of Missing Dairy Data, Ahmed M. Gad Prof, Ahmed Abdelhakim Ahmed Mr, Eman Manaa Prof, Basant Shafik Dr, Sakr Mostafa Prof

Business Administration

Dairy farm records are a crucial component of effective livestock business management. Record analysis allows a farm’s owner to make informed decisions. Incomplete records are less useful for data analysis, so it's important to handle missing values correctly. This study compares different imputation methods for handling missing values in a dataset of dairy records comprising 997 records collected from 234 cows between 2012 and 2022. The dataset was screened against records with missing values and then deleted, resulting in 858 observations from 200 animals. There were missing values in two variables, with a missing percentage of 13.9%: days in milk …

A New Robust Imputation Method For Longitudinal Data With Non-Normal Continuous Outcomes, Ahmed M. Gad Prof, Yasmie A. Mohamed, Nesma M. Daewish Dr, Abdelnaser S. Abdrabou Prof, Wafaa M. Ibrahim Dr

Business Administration

Missing values is very common in longitudinal data and it is the main challenge in analysis of longitudinal data. Missing values have a significant effect on longitudinal data analysis because they lead to loss of information, biased estimates, and misleading results. In practice there is a need for an imputation method to deal with missing values.

Aim: In this study a new robust regression-based imputation method to deal with missing values in longitudinal data is proposed. This method utilizes the modified adaptive linear regression model and does not require the normality of the responses. It is a novel robust imputation …

Extraction Of Fine-Grained Research Methods In The Field Of Information Science, Jiayi Hao, Yuzhuo Wang, Chengzhi Zhang

Journal of Scientific Information Research

[Purpose/significance]Research methods in information science are one of the critical research directions in this field. Constructing a fine-grained research method corpus and extracting research method entities can help scholars quickly understand the research methods in this field, explore the evolution of methods and their future development trends, and lay the foundation for the service and application of the research method corpus in the subsequent digital wave. [Method/process]Firstly, based on academic articles published in the Journal of the China Society for Scientific and Technical Information from 2000 to 2023, this study randomly selected 50 articles and manually annotated the research methodology …

Identifying Research Paths Based On Main Path Analysis: A Case Study Of Knowledge Graphs, Ruibin Wei, Yidan Wang, Yan Xu

Journal of Scientific Information Research

[Purpose/significance]The main path analysis of citation networks can be used to identify important literature in specific fields and can achieve the extraction of mainstream research threads. This paper will use the main path analysis method to analyze the research path of knowledge graphs and sort out the context of their research development. [Method/process]This paper firstly obtains research papers in the field of knowledge graphs from the Web of Science platform, then uses the HistCite software to generate a direct citation network of the literature, and then imports the data into Pajek to generate multiple main paths of the dataset, and …

Association Between Lifetime Interpersonal Violence And Post– Covid-19 Condition Among Women In Kentucky, 2020-2022, Ayşe Güler, Heather M. Bush, Katie Schill, Nurlan Kussainov, Ann L. Coker

Biostatistics Faculty Publications

Objective: The COVID-19 pandemic increased the risk of interpersonal violence. We investigated the association between lifetime interpersonal violence experience and risk of post–COVID-19 condition (the persistence of symptoms of COVID-19 and severity of health problems associated with COVID-19 that last a few weeks, months, or years) among women with lifetime interpersonal violence experience.   Methods: Women participants aged ≥18 years in Kentucky’s Wellness, Health & You—COVID-19 study completed online quantitative surveys about the impacts of the pandemic, developing COVID-19, and symptoms of post–COVID-19 condition. We conducted cross-sectional analyses estimating rate ratios of developing COVID-19 and symptoms of post–COVID-19 condition during the …

Elements Of Statistics, Paul Flesher, Jeffrey Sadler, Jonathan Rehmert, Lanee Young

All Open Educational Resources

This undergraduate text, intended for an introductory statistics course, motivates and develops basic inferential statistics. The first chapter establishes the necessity of inferential statistics, lays the basis of a scientific worldview, and finishes by introducing sampling methods and variables. Subsequent chapters develop visualizations, descriptive statistics, probability, and random variables. Sampling distributions are treated in the fifth chapter which is immediately followed by the development of confidence intervals and hypothesis testing. The text concludes with a treatment of linear regression. The use of Excel is emphasized throughout. The text is best viewed online through LibreTexts.

Temporal Network Analysis Of Comorbidities Among People With Hiv In South Carolina, Yunqing Ma, Matthew Lohman Ph.D., Monique J. Brown Ph.D., Mph, Yichen Li, Xiaoming Li Ph.D., Bankole Olatosi Ph.D., Jiajia Zhang Ph.D.

Faculty Publications

Introduction: People with HIV experience a high rate of comorbidities that can complicate their health outcomes. Understanding the prevalence, interrelationships and temporal development of these comorbidities is crucial for improving health management and quality of life for people with HIV. Methods: We used a population-based cohort extracted from statewide electronic health record (EHR) data in South Carolina (SC), including 18 649 people with HIV who survived at least 1 year after HIV diagnosis between January 1, 2005, and December 31, 2020. Comorbidities and organ systems were classified using ICD-10 codes. Network analysis was performed to assess the closeness centrality among …

‘Waves Of Imagination’ Unconditional Spectogram Diffusion Using Diffusion Architecture., Rahul Vanukuri

Computer Science and Engineering Theses - Archive

The swift evolution of wireless communication technologies,particularly in the field of rf signals or in CBRS bands,demands increasingly sophisticated signal processing techniques to ensure efficient transmission, reception, and spectrum management.Traditional approaches to signal generation and reconstruction, although effective in controlled environments, often struggle to cope with the challenges presented by real-world noisy conditions, hardware constraints, and limited access to large-scale datasets. In response to these limitations, this thesis explores the application of diffusion models—a class of generative models known for their ability to produce high-fidelity samples—to the domain of spectrogram generation for communication signals.

Different from conventional strategies to simulate …

Understanding The Challenges And Satisfaction Among The Medical Professional Preceptors, Cornelia Ko

West Chester University Doctoral Projects

This study evaluates the Penn Medicine Family Medicine Clerkship program through a mixed-methods approach, exploring the challenges and factors influencing the satisfaction of medical professional preceptors when precepting students. A quantitative survey with 54 preceptor participants reveals the top challenges preceptors face: workload, conflicts with patient care, and limited clinical space. Despite this, there is strong support for the program, with 89% of preceptors enjoying the role and 85% expressing willingness to continue to precept for the next three years. Finally, data analysis shows no statistical correlation between preceptor motivation with gender, ethnicity, age group, years of serving, or total …

Highest Risk Density Region For The Communication Of The Impact Of A Treatment Covariate On The Time-To-Event Distribution, Giacomo Biganzoli, Giuseppe Marano Phd, Patrizia Boracchi Phd

COBRA Preprint Series

Analysis of time-to-event (TTE) data is central to clinical research, yet conventional summary measures like the hazard ratio (HR) and restricted mean survival time (RMST) present significant challenges. The HR is often misinterpreted, and its validity depends on the frequently violated proportional hazards assumption, while the RMST is highly sensitive to the choice of time horizon. This paper introduces two novel, assumption-free estimands to address these limitations: the Highest Risk Density Region (HRDR) and the Highest Net Risk Difference Region (HNRDR).

The HRDR identifies the narrowest time interval containing a pre-specified probability mass of events, directly answering the clinical question: …

การเปรียบเทียบการทำนายผลการแข่งขันฟุตบอลโดยใช้การเรียนรู้เชิงลึกจากการผสานข้อมูลทางสถิติและแผนที่ความร้อนของผู้เล่น, ภานุพันธุ์ พันธุ์พฤกษ์

Chulalongkorn University Theses and Dissertations (Chula ETD)

งานวิจัยนี้มีวัตถุประสงค์เพื่อเปรียบเทียบประสิทธิภาพของแบบจำลอง Multilayer Perceptron (MLP), Convolutional Neural Networks (CNN) และ Multimodal Neural Network (MMNN) ในการทำนายผลการแข่งขันฟุตบอลพรีเมียร์ลีกอังกฤษ โดยใช้ข้อมูลทางสถิติ ข้อมูลแผนที่ความร้อนของผู้เล่น และการผสานข้อมูลทั้งสองรูปแบบ ผ่านแนวทาง Early Fusion, Late Fusion และการแปลง Heatmap เป็นข้อมูลเชิงตัวเลขแบบแบ่งพื้นที่สนาม ผลการวิจัยพบว่า แบบจำลอง MMNN ที่ใช้แนวทาง Late Fusion ให้ค่าความแม่นยำที่ 55.68% ซึ่งสูงกว่า Early Fusion ที่ได้ความแม่นยำ 53.45% และ CNN ที่ใช้ข้อมูลทางสถิติร่วมกับข้อมูลตำแหน่งเชิงพื้นที่จากแผนที่ความร้อนที่แบ่งสนามออกเป็น 4 ส่วน มีค่าความแม่นยำสูงสุดที่ 63.50% และให้คะแนน Precision, Recall และ F1 Score สูงกว่าแบบจำลองอื่น ๆ นอกจากนี้ ยังได้วิเคราะห์ความสำคัญของฟีเจอร์ด้วยเทคนิค SHAP ซึ่งชี้ให้เห็นว่าฟีเจอร์ที่เกี่ยวข้องกับเกมรับของทีมเหย้า เช่น การบล็อก การเคลียร์บอล และการดวลลูกกลางอากาศ มีผลต่อการทำนายมากที่สุด ขณะที่ตำแหน่งของผู้เล่นทีมเยือนมีอิทธิพลน้อยที่สุด

Bayesian Merged Utilization Of Grappa And Sense (Bmugs) For In-Plane Accelerated Reconstruction Increases Fmri Detection Power, Chase J. Sakitis, Daniel B. Rowe

Mathematical and Statistical Science Faculty Research and Publications

In fMRI, capturing brain activity during a task is dependent on how quickly the k-space arrays for each volume image are obtained. Acquiring the full k-space arrays can take a considerable amount of time. Under-sampling k-space reduces the acquisition time, but results in aliased, or “folded,” images after applying the inverse Fourier transform (IFT). GeneRalized Autocalibrating Partial Parallel Acquisition (GRAPPA) and SENSitivity Encoding (SENSE) are parallel imaging techniques that yield reconstructed images from subsampled arrays of k-space. With GRAPPA operating in the spatial frequency domain and SENSE in image space, these techniques have been separate but can …

Analyzing Factors Influencing Employee Turnover In Tech Companies: A Predictive Modeling Approach, Shinjon Ghosh

Theses and Dissertations

Employee turnover poses substantial challenges for technology firms, and understanding its key drivers through predictive modeling is essential for developing effective retention strategies. This study investigates factors influencing employee turnover in technology companies by implementing a predictive modeling approach on the IBM HR Analytics Employee Attrition dataset. The research aims were identifying key factors contributing to employee attrition, developing predictive models to forecast turnover risk, and analyzing interactions among significant predictors. By examining a range of features, the results highlight significant variables (Over Time, Monthly Income, Marital Status, etc.) of attrition and offer actionable insights for developing targeted employee retention …

Implementation Of Project-Based Instruction In Statistics Higher Education Courses, Jennifer A. Daddysman

Theses and Dissertations--Education Sciences

Project-based instruction is an instructional methodology based in constructivism that aligns with recommendations from the American Statistical Association for teaching statistics, including incorporating real-world data and examples, collaboration, and scaffolding (GAISE College Report ASA Revision Committee, 2016; GAISE Steering Committee, 2024; Krajcik & Blumenfeld, 2005; Tishkovskaya & Lancaster, 2012). This study examined current use of project-based instruction in United States higher education courses in statistics and the supports for and challenges to implementing this type of instruction.

This study used qualitative research methodology in the integrative pedagogy and diffusion of innovation frameworks. Voluntary response surveys were distributed via two sections …

Understanding The Impact Of The Medicaid Expansion On Hospital Length Of Stay And Emergency Department Use In Kentucky, Cameron Bushling

Theses and Dissertations--Epidemiology and Biostatistics

Three related analyses were performed to try to understand the immediate effects of Medicaid expansion on hospital systems in Kentucky. Medicaid expansion led to a large, sudden increase in the number of individuals eligible for various health care needs. This sudden increase in demand for health services lead to initial hypotheses regarding hospitals’ ability to handle this demand. The first analysis examined hospital average length of stay (LOS) using a linear mixed-effects model. Length of stay was modeled longitudinally between 2013 and 2015 to determine if significant changes in the slope of LOS could be detected after expansion (2014). Separate …

Changes In Cancer Diagnosis And Survival In The United States During The Covid-19 Pandemic, Justin T. Burus

Theses and Dissertations--Epidemiology and Biostatistics

The COVID-19 Pandemic led to global societal disruptions as political leaders and public health authorities attempted to control the spread of the newly discovered SARS- CoV-2 virus. While these measures were designed to lessen morbidity and mortality from a novel pathogen, their impact was also felt in many other, often unintended ways. The purpose of this dissertation is to use cancer surveillance research methods to examine the association between COVID-19 Pandemic-related disruptions and changes in the normal diagnosis and care of cancer in the United States.

The first two studies of this dissertation analyzed reductions in cancer diagnoses in the …