Data Science Commons^™

Machine Learning And Geostatistical Approaches For Discovery Of Weather And Climate Events Related To El Niño Phenomena, Sachi Perera

Computational and Data Sciences (PhD) Dissertations

El Nino and La Nina are worldwide environmental phenomena brought about by repetitive changes in the water temperature of the Pacific Ocean. Even though the El-Nino impact focuses on a smaller area in the Pacific Ocean near the Equator, these developments have global repercussions, where temperature and precipitation are influenced across the globe, causing droughts and floods simultaneously. In this dissertation, we first derived a drought vulnerability index for the Nile basin, identifying regions with high and low drought risk under ENSO conditions. Next, we evaluated the coherence and periodicity of the ENSO signal to detect its implications on MENA …

The Quantitative Analysis And Visualization Of Nfl Passing Routes, Sandeep Chitturi

Computer Science and Computer Engineering Undergraduate Honors Theses

The strategic planning of offensive passing plays in the NFL incorporates numerous variables, including defensive coverages, player positioning, historical data, etc. This project develops an application using an analytical framework and an interactive model to simulate and visualize an NFL offense's passing strategy under varying conditions. Using R-programming and data management, the model dynamically represents potential passing routes in response to different defensive schemes. The system architecture integrates data from historical NFL league years to generate quantified route scores through designed mathematical equations. This allows for the prediction of potential passing routes for offensive skill players in response to the …

Sequential Optimization For Stressor-Informed Test Planning Through Integration Of Experimental And Simulated Data, Jacob Brecheisen

Data Science Undergraduate Honors Theses

This technical report details an innovative approach in reliability engineering aimed at maximizing system durability through a synergistic use of physical experimentation and computer-based modeling. Our methodology explores the efficient design and analysis of computer experiments and physical tests to facilitate accelerated reliability growth, while leveraging a sequential integration of data from these two distinct sources: costly physical experiments, characterized by random errors, and inexpensive computer simulations, marked by inherent systematic errors. The key innovation lies in the adoption of a closed-loop design and analysis method. This method begins by identifying a viable subset of important environmental stressors—such as temperature, …

Murmurations And Root Numbers, Alexey Pozdnyakov

University Scholar Projects

We report on a machine learning investigation of large datasets of elliptic curves and L-functions. This leads to the discovery of murmurations, an unexpected correlation between the root numbers and Dirichlet coefficients of L-functions. We provide a formal definition of murmurations, describe the connection with 1-level density, and provide three examples for which the murmuration phenomenon has been rigorously proven. Using our understanding of murmurations, we then build new machine learning models in search of a polynomial time algorithm for predicting root numbers. Based on our models and several heuristic arguments, we conclude that it is unlikely for …

Global To Glocal: A Confluence Of Data Science And Earth Observations In The Advancement Of The Sdgs, Rejoice Thomas

Computational and Data Sciences (PhD) Dissertations

The United Nations' (UN) Sustainable Development Goals (SDGs), part of Agenda 2030, comprise 17 interconnected goals and 169 actionable targets, providing an effective framework for addressing diverse issues ranging from individual challenges such as poverty, hunger, and health to broader corporate and global challenges like climate change and equality. Among these interconnected SDGs, this dissertation focuses on the role of climate and infrastructure in global and local sustainability. To this end, earth observations have been conducted utilizing data science techniques to advance these SDGs. For this dissertation, the author has conducted earth studies serving the following SDGs:

1. SDG 3 (Good …

A Comprehensive Analysis Of Training Induced Heat-Related Injuries At Fort Moore, Anthony Beger

Data Science Undergraduate Honors Theses

Heat related injuries are a significant problem for the United States Armed Forces. There were over 11,000 confirmed cases of heat-related illnesses that were diagnosed at more than 230 military installations from 2018-2022. These injuries are primarily due to hyperthermia (i.e., abnormally high body temperature) resulting from extreme environmental temperatures, high humidity, medications, or excessive physical work or exercise. Fort Moore has the most heat related injuries of any installation in the U.S. Department of Defense since it is home to one of the largest U. S. Army training posts with most training involving intensive outdoor activity in high heat …

Implementation Of Explainable Ai For Bearing Fault Classification, Mohammad Mundiwala

Honors Scholar Theses

It is difficult to overstate the impact of artificial intelligence (AI) over the past decade. The rapid expansion of machine learning has stimulated a race to deploy AI in all facets of life, one such domain being machine health monitoring. There is no doubt that machine learning excels in prediction accuracy, but oftentimes, these models are cryptic and fail to provide valuable insight into their decisions. This paper presents an overview of a neural network and what it means to learn. Next, two distinct Explainable AI (XAI) techniques will be presented: Gradient Class Activation Mapping and SimplEx . Finally, these …

Advancement Of Iterative Optimization Technology Algorithms Toward Calibration-Free Process Analytical Technology Applications, Adam Rish

Electronic Theses and Dissertations

The expansion of spectroscopic process analytical technology (PAT) tools within the pharmaceutical industry has the potential to elevate the current state-of-the-art of pharmaceutical manufacturing by offering opportunities for reduced quality testing times, enhanced process control, and greater production flexibility. Spectroscopic PAT tools are dependent on multivariate models to extract the relevant information from the spectral outputs. However, there is a substantial calibration burden for developing and maintaining these multivariate models that discourages the application of PAT, despite the encouragement from regulators. This has led to an interest in calibration-free methods such as iterative optimization technology (IOT) for spectroscopic PAT that …

Spatiotemporal Negative Inventory Outlier Decomposition For Supply Chain Applications In Consumer-Packaged Goods (Cpg), Hayden Mcdonald

Data Science Undergraduate Honors Theses

Coca-Cola is a popular soft drink brand with sales occurring in every Walmart store across the world, which generates large quantities of data and requires a robust supply chain system. However, the company does not currently have a sophisticated, automated, and/or prescriptive system for detecting where, when, and why inventory outages occur and applying preventative measures to avoid loss of revenue from the absence of inventory on store shelves. This thesis proposes and applies a novel, prescriptive system for this purpose. An inventory outage can be seen as a ‘negative’ statistical outlier in a time series of inventory for an …

The Importance Of Data Preparation In A Data Science Problem, Sophia Beard

Data Science Undergraduate Honors Theses

This study is going to be based on an inventory outlier automation data science problem that is being solved to identify and prescribe inventory level outliers to help keep shelves stocked in terms of beverages. The objective of this paper will address why it is so important to understand the data that is involved in a particular data science problem and how planning ahead ensures a successful outcome in the data science world. In this data science project, Spatiotemporal Outlier Analysis for Inventory Intervention Automation, it was crucial for the team to understand, research, and visualize the data we were …

Employing Natural Language Processing To Link Customer Survey Feedback With Net Promoter Scores, Gerardo Moreno

Data Science Undergraduate Honors Theses

This project leverages Natural Language Processing (NLP) to analyze customer feedback from Sam’s Club, aiming to pinpoint key factors influencing Net Promoter Score (NPS). Using sentiment analysis, bigram, and trigram techniques, the project analyses textual data to identify underlying themes and patterns that affect customer satisfaction. These analyses reveal actionable insights into customer preferences and pain points, facilitating a deeper understanding of what drives customer satisfaction in retail environments. By correlating these findings with NPS, this paper details strategies to enhance customer experiences at Sam’s Club, ultimately aiming to improve both satisfaction levels and NPS.

A Spatiotemporal Analysis Of Violent Crime In Little Rock, Arkansas From 1999-2022, Nicole Rogers

Data Science Undergraduate Honors Theses

Little Rock, Arkansas is not only the capital and largest city in Arkansas, but it has one of the highest crime rates amongst cities with over 100,000 people in the country. According to the US Census in 2020, Little Rock had a population of 202,591. In the same year, Little Rock Police Department recorded 3,567 cases of violent crime, leading to a violent crime rate of 1,805 violent crime occurrences per 100,000 people. For perspective, Chicago’s violent crime rate was approximately half of that of Little Rock during the same time period. Crime, like other social phenomena is unevenly distributed …

Low-Resource Icd Coding Of Discharge Summaries, Ashton Williamson

All Theses

Medical coding is the process by which standardized medical codes are assigned to patient health records. This is a complex and challenging task that typically requires an expert human coder to review health records and assign codes from a classification system based on a standard set of rules. Considering the downstream use of these codes in statistical analysis, billing, and patient care, improving the accuracy and efficiency of the medical coding process through automation could have a far-reaching impact on the healthcare domain. Since health records typically consist of a large proportion of free-text documents, this problem has traditionally been …

Comparative Predictive Analysis Of Stock Performance In The Tech Sector, Asaad Sendi

LSU New Orleans Theses and Dissertations

This study compares the performance of deep learning models, including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Transformer, in predicting stock prices across five companies (AAPL, CSCO, META, MSFT, and TSLA) from July 2019 to July 2023. Key findings reveal that GRU models generally exhibit the lowest Mean Absolute Error (MAE), indicating higher precision, particularly notable for CSCO with a remarkably low MAE. While LSTM models often show slightly higher MAE values, they outperform Transformer models in capturing broader trends and variance in stock prices, as evidenced by higher R-squared (R2) values. Transformer models generally exhibit higher MAE …

Igwas: Image-Based Genome-Wide Association Of Self-Supervised Deep Phenotyping Of Retina Fundus Images, Ziqian Xie, Tao Zhang, Sangbae Kim, Jiaxiong Lu, Wanheng Zhang, Cheng-Hui Lin, Man-Ru Wu, Alexander Davis, Roomasa Channa, Luca Giancardo, Han Chen, Sui Wang, Rui Chen, Degui Zhi

Faculty, Staff and Student Publications

Existing imaging genetics studies have been mostly limited in scope by using imaging-derived phenotypes defined by human experts. Here, leveraging new breakthroughs in self-supervised deep representation learning, we propose a new approach, image-based genome-wide association study (iGWAS), for identifying genetic factors associated with phenotypes discovered from medical images using contrastive learning. Using retinal fundus photos, our model extracts a 128-dimensional vector representing features of the retina as phenotypes. After training the model on 40,000 images from the EyePACS dataset, we generated phenotypes from 130,329 images of 65,629 British White participants in the UK Biobank. We conducted GWAS on these phenotypes …

Improvements In Appropriate Placement Of Dental Sealants After Implementation Of A Clinical Decision Support System, Joanna Mullins, Ryan Brandon, Nicholas Skourtes, Elsbeth Kalenderian, Muhammad Walji

Faculty, Staff and Student Publications

BACKGROUND: Dental sealants are effective for the prevention of caries in children at elevated risk levels, and increasing the proportion of children and adolescents who have dental sealants on 1 or more molars is a Healthy People 2030 objective. Electronic health record (EHR)-based clinical decision support systems (CDSSs) have the ability to improve patient care. A dental quality measure related to dental sealant placement for children at elevated risk of caries was targeted for improvement using a CDSS.

METHODS: A validated dental quality measure was adapted to assess a patient's need for dental sealant placement. A CDSS was implemented to …

Development And Validation Of A Rule-Based Algorithm To Identify Periodontal Diagnosis Using Structured Electronic Health Record Data, Bunmi Tokede, Ryan Brandon, Chun-Teh Lee, Guo-Hao Lin, Joel White, Alfa Yansane, Xiaoqian Jiang, Elsbeth Kalenderian, Muhammad Walji

Faculty, Staff and Student Publications

AIM: To develop and validate an automated electronic health record (EHR)-based algorithm to suggest a periodontal diagnosis based on the 2017 World Workshop on the Classification of Periodontal Diseases and Conditions.

MATERIALS AND METHODS: Using material published from the 2017 World Workshop, a tool was iteratively developed to suggest a periodontal diagnosis based on clinical data within the EHR. Pertinent clinical data included clinical attachment level (CAL), gingival margin to cemento-enamel junction distance, probing depth, furcation involvement (if present) and mobility. Chart reviews were conducted to confirm the algorithm's ability to accurately extract clinical data from the EHR, and then …

Predicting True Attributes Of Retailer Data, Abby Willard

Data Science Undergraduate Honors Theses

In the rapidly evolving landscape of consumer-packaged goods (CPG) retail, understanding the true values of various factors influencing sales performance is paramount for strategic decision-making and effective resource allocation. In ensuring accuracy of data points, the CatBoost model is utilized, a state-of-the-art gradient boosting technique, to predict the true attribution values of datasets sourced from CPG industry retailers.

By leveraging CatBoost’s inherent capabilities to handle categorical data and its robustness against overfitting, the models are optimized to accurately predict the true attribution values for various items. The performance of the CatBoost models is evaluated through rigorous cross-validation techniques and compared …

Concurrent Processing Of Retail Data In Python To Optimize Runtime, Bobby Slavin

Data Science Undergraduate Honors Theses

This thesis explores the application of multiprocessing and multithreading techniques in Python to optimize runtime efficiency on the analysis of retail data. As the retail data processed by a program increases, so does the runtime of the program. If you are performing this processing using only a single core, even a gigabyte of data can potentially take upwards to half an hour to finish processing, while larger datasets of 100 GB or more could take days, heavily limiting the amount of retail data that can be processed in a reasonable amount of time. By employing multithreading and multiprocessing architectures in …

The Importance Of Text Representation For Neural Networks Through Natural Language Processing Techniques, William Parsley

Data Science Undergraduate Honors Theses

Text representation is a fundamental aspect of natural language processing (NLP) when it comes to the performance of neural networks. Free-form text fields are being utilized in more and more industries. Anything from a description of an item on a web store to tracking service events to military-grade aircraft is being collected in free-form text. The goal of the thesis is to highlight best practices and discuss trends in data to prepare text for a neural network. It will demonstrate various techniques for representing free-form text in the context of neural networks, focusing on data preparation decisions, embedding techniques, and …

Examining The Impact Of Customer Rfp Characteristics On Award Compliance, Laasya Ravipati

Data Science Undergraduate Honors Theses

In the context of intermodal transportation, understanding the dynamics of award compliance holds significant importance for operational efficiency and strategic decision-making. Award compliance refers to the percentage of awarded freight volume that is realized, indicating the extent to which contractual agreements are fulfilled. This analysis delves into the intricate relationship between customer characteristics and award compliance, aiming to provide valuable insights into the variability and predictability of compliance rates. By analyzing Request for Pricing (RFP) data and primary awarded freight volumes, the study seeks to address the need for more accurate volume estimations, crucial for sales planning, revenue projections, and …

Examining Award Compliance To Inform Resource Allocation, Jacob Haarala

Data Science Undergraduate Honors Theses

This project focuses on JB Hunt Transport Inc's intermodal business unit (JBI) by focusing on the challenges associated with Published Pricing and Contractual Pricing. The primary issue revolves around the variance between the awarded freight volumes in Requests for Pricing (RFPs) and the actual volumes realized when the freight is shipped. This discrepancy poses challenges for effective sales planning, revenue goals, and optimal freight network management within JBI. Reporting tools, such as PowerBI, are currently used by JBI to provide insights into award compliance on a weekly basis. However, our goal with this project was to provide a deeper understanding …

The Effects Of Using Machine Translators On The Performance Of Second Language Learners, Kasey Myer

University Honors College

With the rise of technology has also come the development of various online language translators and artificial intelligence that are often utilized by individuals learning a second language. However, there is a wide range of quality between the different machine translation tools, and many people tend to be under the impression that it is inferior to the quality of interpretations provided by human translators. This paper considers the positives and negatives of machine translation as a tool for second language learning. Variations between the input and output languages on a grammatical and cultural level are analyzed. Machine translation is compared …

Per- And Polyfluoroalkyl Substance Exposure Risks In Us Carceral Facilities, 2022, Lindsay Poirier, Derrick Salvatore, Phil Brown, Alissa Cordner, Kira Mok, Nicholas Shapiro

Statistical and Data Sciences: Faculty Publications

Objectives. To assess the US incarcerated population’s risk of exposure to per- and polyfluoroalkyl substances (PFASs). Methods. We assessed how many of the 6118 US carceral facilities were located in the same hydrologic unit code watershed boundaries as known or likely locations of PFAS contamination. We conducted geospatial analyses on data aggregated from Environmental Protection Agency databases and a PFAS site tracker in 2022 to model the hydrologically feasible known and presumptive PFAS contamination sites for nearly 2 million incarcerated people. Results. Findings indicate that 5% (∼310) of US carceral facilities have at least 1 known source of PFAS contamination …

Leveraging Logical Definitions And Lexical Features To Detect Missing Is-A Relations In Biomedical Terminologies, Rashmie Abeysinghe, Fengbo Zheng, Jay Shi, Samden D Lhatoo, Licong Cui

Faculty, Staff and Student Publications

Biomedical terminologies play a vital role in managing biomedical data. Missing IS-A relations in a biomedical terminology could be detrimental to its downstream usages. In this paper, we investigate an approach combining logical definitions and lexical features to discover missing IS-A relations in two biomedical terminologies: SNOMED CT and the National Cancer Institute (NCI) thesaurus. The method is applied to unrelated concept-pairs within non-lattice subgraphs: graph fragments within a terminology likely to contain various inconsistencies. Our approach first compares whether the logical definition of a concept is more general than that of the other concept. Then, we check whether the …

Natural Language Processing Of Clinical Notes Enables Early Inborn Error Of Immunity Risk Ascertainment, Kirk Roberts, Aaron T Chin, Klaus Loewy, Lisa Pompeii, Harold Shin, Nicholas L Rider

Faculty, Staff and Student Publications

BACKGROUND: There are now approximately 450 discrete inborn errors of immunity (IEI) described; however, diagnostic rates remain suboptimal. Use of structured health record data has proven useful for patient detection but may be augmented by natural language processing (NLP). Here we present a machine learning model that can distinguish patients from controls significantly in advance of ultimate diagnosis date.

OBJECTIVE: We sought to create an NLP machine learning algorithm that could identify IEI patients early during the disease course and shorten the diagnostic odyssey.

METHODS: Our approach involved extracting a large corpus of IEI patient clinical-note text from a major …

Montelukast As A Repurposable Additive Drug For Standard-Efficacy Multiple Sclerosis Treatment: Emulating Clinical Trials With Retrospective Administrative Health Claims Data, Astrid M Manuel, Assaf Gottlieb, Leorah Freeman, Zhongming Zhao

Faculty, Staff and Student Publications

BACKGROUND: Effective and safe treatment options for multiple sclerosis (MS) are still needed. Montelukast, a leukotriene receptor antagonist (LTRA) currently indicated for asthma or allergic rhinitis, may provide an additional therapeutic approach.

OBJECTIVE: The study aimed to evaluate the effects of montelukast on the relapses of people with MS (pwMS).

METHODS: In this retrospective case-control study, two independent longitudinal claims datasets were used to emulate randomized clinical trials (RCTs). We identified pwMS aged 18-65 years, on MS disease-modifying therapies concomitantly, in de-identified claims from Optum's Clinformatics

RESULTS: pwMS treated with montelukast demonstrated a statistically significant 23.6% reduction in relapses compared …

Age At Lung Cancer Diagnosis In Females Versus Males Who Never Smoke By Race And Ethnicity, Batel Blechter, Jason Y Y Wong, Li-Hsin Chien, Kouya Shiraishi, Xiao-Ou Shu, Qiuyin Cai, Wei Zheng, Bu-Tian Ji, Wei Hu, Mohammad L Rahman, Hsin-Fang Jiang, Fang-Yu Tsai, Wen-Yi Huang, Yu-Tang Gao, Xijing Han, Mark D Steinwandel, Gong Yang, Yihe G Daida, Su-Ying Liang, Scarlett L Gomez, Mindy C Derouen, W Ryan Diver, Ananya G Reddy, Alpa V Patel, Loïc Le Marchand, Christopher Haiman, Takashi Kohno, Iona Cheng, I-Shou Chang, Chao Agnes Hsiung, Nathaniel Rothman, Qing Lan

Faculty, Staff and Student Publications

BACKGROUND: We characterized age at diagnosis and estimated sex differences for lung cancer and its histological subtypes among individuals who never smoke.

METHODS: We analyzed the distribution of age at lung cancer diagnosis in 33,793 individuals across 8 cohort studies and two national registries from East Asia, the United States (US) and the United Kingdom (UK). Student's t-tests were used to assess the study population differences (Δ years) in age at diagnosis comparing females and males who never smoke across subgroups defined by race/ethnicity, geographic location, and histological subtypes.

RESULTS: We found that among Chinese individuals diagnosed with lung cancer …

Plumbing The Depths Of The Shallow End: Exploring Persistent Homology Using Small Data, R. Anne Flynn

All NMU Master's Theses

Persistent homology is a prominent tool in topological data analysis. This thesis is designed to be an introduction and guide to a beginner in persistent homology. This comprehensive overview discusses the math used behind it, the code needed to apply it, and its current place in the field. We explain and demonstrate the algebraic topology which fuels persistent homology. Homotopies inspire homology groups, which are able to determine how many holes a shape has. By visualizing data as a shape, persistent homology determines what type of holes are present.

We demonstrate this by using the package TDA in the manipulation …

Analyzing Information Diffusion In Social Media Networks, Amin Riazi

Masters Theses and Doctoral Dissertations

Social media is a dynamic platform where a wide range of information is shared, including both true and false content, and it involves interactions between human users and social bots. This study investigates information diffusion patterns on X (formerly Twitter) by analyzing retweet (repost) network topologies. The results reveal distinct behavioral patterns for humans and bots when spreading true and false information, highlighting the need for further examination of their roles in information dissemination. Moreover, this study tackles the challenge of differentiating between broadcast and viral information diffusion on X, acknowledging the possibility of genuine information also being potentially misleading. …