Physical Sciences and Mathematics Commons^™

Algorithmic Monocultures In Hiring, Rishi Bommasani, Sarah H. Bana, Kathleen A. Creel, Dan Saul Jurafsky, Percy Liang

Economics Faculty Articles and Research

Many employers screen job applicants with algorithms built by the same few algorithm vendors. We hypothesize that algorithmic monoculture leads to the same individuals and members of the same racial groups facing rejection. We acquire and analyze a novel dataset of 3 million applicants submitting 4 million applications where all the applications are screened by algorithms built by the same vendor. We find clear racial disparities in applicant outcomes. Of all applications submitted by Asian and Black applicants, 14.74% and 25.87% are submitted to positions that adversely impact Asian and Black applicants, respectively, according to U.S. employment discrimination standards. Individuals …

Learning Trajectories Of Online Batch Selection Methods, Luke Green

Theses and Dissertations

Modern deep neural networks achieve strong performance on large-scale datasets, but often require substantial training time. Online batch selection methods seek to reduce this cost by updating models on informative subsets of each batch rather than on all available examples. Recently introduced methods leverage teacher models and report substantial speedups, particularly in noisy-label settings. However, comparisons are often based on the number of epochs required to reach a target test accuracy, a coarse metric that is sensitive to implementation details and may obscure important differences in learning dynamics. In this thesis, we implement several online batch selection methods in a …

Integration Of Intraoperative Data In Interpretable Machine Learning Models To Predict Postoperative Aki In Noncardiac Surgery Patients, Justin Do, Karan H. Shah, Melissa Xu, Andrew Hyunwoo Kim, Vivaswat Suresh, Nidhir Guggilla, Michael Li, Rishi Kothari

Department of Anesthesiology Faculty Papers

OBJECTIVES: We aimed to (1) quantify changes in discrimination when adding intraoperative data to preoperative data and (2) compare tabular machine learning with feature engineering against a time-aware LSTM-based model.

MATERIALS AND METHODS: Retrospective cohort of 46 204 adults undergoing 57 055 eligible noncardiac surgery in the INSPIRE database. We extracted 38 preoperative and 49 intraoperative variables; acute kidney injury (AKI) was defined by KDIGO serum creatinine criteria and modeled as stage 2/3 postoperative AKI. Models were trained on preoperative-only and combined pre- and intraoperative data. Intraoperative series were summarized using eight statistical features for tabular models or integrated directly …

Analysis And Machine Learning Adaptation Of A Cognitive Model For Human Memory, Trevor Cross, Aihua W. Wood

Faculty Publications

In this paper, we use the Duolingo SLAM dataset to analyze several cognitive models of second language acquisition and develop new approaches for enhanced performance. In particular, we consider the Predictive Performance Equation and some of its underlying power laws. Leveraging insights from machine learning, we develop simple one-feature models as building blocks for combined models that match or in certain cases outperform the existing models at much reduced computational cost. In addition, a neural network with one fully connected hidden layer is constructed that outperforms all other models on sufficiently large datasets.

Data-Driven Electrochemistry Reveals The Impact Of Hydrophobicity On Aptamer Cross-Reactivity, Emily Carroll, Michael A. Pence, Elizabeth Winterholler, Taylor D. Sparks, Shelley D. Minteer

Chemistry Faculty Research & Creative Works

Electrochemical aptamer-based (E-AB) biosensors offer a promising platform for reagentless detection of molecular targets, yet aptamer recognition can be limited by cross-reactivity, particularly for hydrophobic analytes such as steroid hormones. To investigate how cross-reactivity influences E-AB sensor performance, we use automation and machine learning to screen a library of possible interferent molecules against a steroid-binding aptamer, with progesterone serving as a physiologically relevant test case. Here, we develop a label-free E-AB sensor for progesterone detection using a methylene blue-modified aptamer anchored with a hexanethiol linker. We then used an automated electrochemistry platform to perform reproducible and high-throughput characterization of our …

Quantifying Terrain Controls On Satellite-Based Snow Water Equivalent Estimation: A Spatially Explicit Machine Learning Approach, Brant Giovannetti

Geography and the Environment: Graduate Student Capstones

Terrain variables are widely incorporated into machine learning Snow Water Equivalent (SWE) models but are rarely evaluated for their independent contribution relative to spectral predictors. Using a four-tier stepwise Random Forest framework with Harmonized Landsat Sentinel-2 imagery and Airborne Snow Observatory LiDAR ground truth, this study isolates the contribution of elevation, slope, northness, and eastness across Peak and Ablation snowpack regimes in the East Taylor River Watershed, Colorado. During peak snowpack, adding terrain improved R² by 0.214, with elevation alone accounting for 42.8% of model importance. During ablation, full-dataset terrain gains were modest, increasing R² by only 0.036. However, when …

A Systematic Review Of Intrusion Detection Systems For Internet Of Medical Things: Performance, Efficiency, Explainability, And Generalization, Oswald Adohinzin, Youssef Harrath

Research & Publications

The Internet of Medical Things (IoMT) has transformed health care delivery through medical devices, remote patient monitoring, and real-time clinical decision support. However, the proliferation of IoMT devices introduces security vulnerabilities that put patient safety and data privacy at risk. Intrusion Detection Systems (IDS) have emerged as essential components for protecting IoMT networks from cyberattacks. This article presents a systematic review of IoMT-IDS research, analyzing 53 high-quality papers published between 2020 and 2025, identified through database searches spanning 2016–2025 across IEEE Xplore, Springer, ScienceDirect, and ACM Digital Library. We organize the literature through a comprehensive taxonomy spanning classical machine learning …

Developing A Humpback Whale Vocalization Detector Using Machine Learning Models, Lucas Kantorowski

Master's Theses

Humpback whale songs are notoriously complex. Identification of humpback whale song units requires bioacousticians to tediously listen, analyze, and annotate collected sound data. Even sparse data requires listening to the entirety of the collected acoustic data. In this study, three hours of audio containing over one-thousand humpback whale song units was collected in Monterey Bay, California.

Prior studies have seen success using convolutional neural networks by performing image classification on hundreds of hours worth of spectrograms. Our study uses traditional machine learning models, as they are less computationally demanding, and require less data.

We use time splitting and Mel-frequency cepstrum …

A Proactive Food Demand Forecasting-Inventory Management Approach Under Weather Disruptions, Asmaa Seyam, Sujith Samuel Mathew, May El Barachi, Jun Shen

All Works

Effective demand forecasting has become crucial to strengthening system resilience, reducing food waste, and achieving sustainability in food systems. Despite recent advances in leveraging machine learning for food demand forecasting, most existing models remain static and assume stable demand patterns, posing a challenge for adapting to demand changes during disruption events. This paper develops a proactive approach that leverages demand forecasting outputs and weather disruption flags to guide inventory replenishment, ensuring adaptability to varying demand conditions across three weather disruption events while reducing waste. This paper first uses a stacking model to predict next-day demand for a food retailer, leveraging …

Applications Of Prior And Novel Computational Tools In Mental Health Treatment, And Their Potential To Uncover The Explanatory Gap, Ambika Vyas

University Honors Theses

The explanatory gap is a widely discussed concept in scientific and philosophical literature. In neuroscience, the solution to the explanatory gap is highly sought out, but the general consensus is that it is unsolvable. Numerous articles discuss the explanatory gap alongside computational tools and how these tools could aid neuroscientists in uncovering the mental health explanatory gap. However, significant developments in machine learning have been made since 2020, coinciding with the rise in Large Language Models (LLMs). This thesis is a literature review on computational methods, tools, and devices developed and utilized by researchers to improve how mental health disorders …

Multimodal Machine Learning For Soil Burn Severity Mapping Across California Wildfires, Sanjana Checker

Master's Theses

Accurate mapping of soil burn severity (SBS) is critical for post-fire watershed management, erosion risk assessment, and ecological recovery planning, yet traditional field-based approaches remain costly, time-intensive, and spatially limited. This thesis presents a machine learning pipeline for wall-to-wall SBS classification across California wildfires using multi-sensor satellite imagery, terrain derivatives, and bioclimatic covariates. Field-collected SBS observations (n = 2,180) from 52 wildfires occur- ring between 2013 and 2025, sourced from the U.S. Forest Service and CAL FIRE, were used to train and evaluate multiple classification architectures within a Google Earth Engine and Google Cloud-based prediction framework. After upsampling the unburned …

What Makes A Modern Attention Implementation?, Brian H. Slonim

Master's Theses

Since the seminal assertion by Vaswani et al. in 2017 that “Attention Is All You Need,” transformer models have risen to ubiquity due to their ability to learn extremely complex patterns from sequence data, culminating in the unprecedented generative capabilities of large language models. These models’ strength lies in their scale: hundreds of millions (e.g., BERT-LARGE) to billions or trillions of learned parameters. Running inference with these models, let alone training them, would be intractable without significant innovations in the hardware and software that support them. This need has driven an enormous demand for GPU compute and associated software ecosystems, …

Crab: A Novel Clustering Score Using Clustering With Rivals And Buddies For Unsupervised Learning, Allen Choi

Master's Theses

Unsupervised clustering algorithms today are used across a wide variety of fields such as biology, engineering, and industry in order to classify observations into groups where labels are not provided. This can provide important latent information regarding the observations within groups, as well as insight regarding the groups themselves. In order to judge the optimal number of clusters for an unsupervised clustering algorithm, many methods exist such as the Elbow Method and Silhouette Score; however, these methods come with drawbacks and are not necessarily flexible across many unsupervised methods. We present a novel clustering score framework relying on a resampling-based …

Inductive Biases In Field-Level Cosmological Inference From Galaxy Catalogs, James O'Connor Baldwin

Dissertations, Theses, and Capstone Projects

We perform field-level likelihood-free inference of the matter density parameter Ωm from simulated galaxy catalogs using machine learning models with differing inductive biases. Using features extracted from hydrodynamic simulations in the CAMELS suite, we investigate how both observable choice and model architecture govern the extraction of cosmological information. We consider galaxy positions and line-of-sight peculiar velocities, both separately and in combination, and compare permutation-invariant Deep Sets, implemented with either standard multilayer perceptrons (MLPs) or Kolmogorov–Arnold Networks (KANs), to graph neural networks (GNNs) implemented with MLPs, which explicitly encode spatial relations. We evaluate inference performance under both in-distribution and out-of-distribution (OOD) …

Mapping Public Engagement With Environmental Finance In The U.S: Spatiotemporal Insights From Google Trends On Public Engagement And Policy Equity In The U.S, Hilda Afeku-Amenyo, Charles Knoble, Pankaj Lal

Department of Earth and Environmental Studies Faculty Scholarship and Creative Works

This study analyzes U.S. public engagement with environmental finance from 2015–2023 using Google Trends data for four key terms: “green bonds,” “sustainable finance,” “climate finance,” and “ESG.” Our analysis reveals a notable rise in engagement with environmental finance topics, with Environmental, Social, and Governance (ESG) investing generating the highest levels of interest. Geographically, the District of Columbia exhibited disproportionate search activity relative to other regions. While public awareness of environmental finance has grown, we highlight disparities in knowledge and accessibility of information. These findings carry significant policy implications, particularly in ensuring equitable access to sustainable financial instruments. We underscore the …

Bayesian Optimization Of Laser Wakefield Acceleration Via Spectral Pulse Shaping, B. Z. Djordjević, C. Benedetti, Asa D. Mcnaughton, R. Lehe, H.-E. Tsai, S. C. Wilks, B. A. Reagan, G. J. Williams, J. Van Tilborg, C. B. Schroeder

Physics Faculty Publications and Presentations

In this paper, we investigate the effect of spectral pulse shaping of the laser driver on the performance of channel-guided, laser–plasma accelerators. The study was carried out with the assistance of Bayesian optimization using particle-in-cell simulations. We used a realistic plasma profile based on a novel optical-field-ionized channel technique with ionization injection and low, on-axis plasma densities to maximize the energy gain of the electron bunch trailing the laser. Spectral shaping allows us to modify the temporal profile of the laser driver while keeping the laser energy constant, affecting the acceleration and injection processes. In addition, we consider how modifying …

A Hybrid Approach To Phishing Email Detection: Leveraging Machine Learning And Large Language Models, Hessa Shamal Biri

Theses

Phishing attacks have reached a new level of sophistication through the deployment of large language models by attackers. The current AI-generated threats defeat existing detection systems which base their operation on past data. The thesis presents a hybrid system for phishing email detection which combines real email data with synthetic LLM-created samples to enhance traditional machine learning classifiers performance. The study created a hybrid dataset of 20,627 emails by combining 18,631 real messages from the Kaggle Email Classification Dataset with 1,996 synthetic emails. The synthetic emails were generated using four large language models LLaMA-3, Falcon, LLaVA, and Mistral to capture …

Principles Of Privacy And Security In Artificial Intelligence And Applications, Khang Tran

Dissertations

Modern artificial intelligence (AI) systems have transformed critical domains such as healthcare, software engineering, finance, and the legal system. Despite their broad impact, concerns about trustworthiness, especially regarding privacy and security, remain major obstacles to wider adoption. Addressing these concerns requires both a systematic understanding of the privacy and security risks inherent in AI systems and the development of principled foundations for trustworthy AI that safeguard client privacy and security. This goal is particularly challenging because of the complexity of modern large-scale AI systems, the trade-offs between privacy and model utility, and the need to simultaneously ensure other important properties …

Differential-Geometric Methods For Neural Signed Distance Fields: Parameterized Surface Extraction And Curvature Regularization For Cad Models, Haotian Yin

Dissertations

Neural signed distance fields have emerged as a powerful framework for representing three-dimensional geometry through continuous and differentiable neural functions. Their flexibility, resolution independence, and compatibility with gradient-based optimization make them especially attractive for surface reconstruction and geometric learning. However, despite these advantages, two fundamental challenges remain for engineering-grade applications. First, higher-order geometric properties such as curvature are difficult to model reliably during training and often require computationally expensive second-order differentiation. Second, while neural signed distance fields provide implicit surface representations, they do not directly yield a globally consistent forward map or parameterization for downstream geometric processing.

This dissertation addresses …

A Generative Ai-Driven Computational Framework For Industry-Scale Discovery Of Novel Battery Materials, Joy Datta

Dissertations

The growing demand for sustainable, high-energy-density electrochemical storage has motivated the exploration of multivalent-ion batteries based on earth-abundant elements such as aluminum, calcium, magnesium, and zinc. While multivalent charge carriers offer higher theoretical energy density than lithium, their practical deployment is hindered by sluggish ion transport, strong ion-host interactions, and structural degradation of electrode materials. Identifying host materials that can reversibly accommodate multivalent ions while maintaining structural integrity remains a fundamental challenge. The dissertation develops a scalable, end-to-end computational framework that integrates density functional theory (DFT), machine learning (ML), and generative artificial intelligence (GenAI) to accelerate the discovery of next-generation …

Toward Learning-Based Reconstruction And Part Decomposition Of Man-Made 3d Geometry: Neural Implicit Representations And Scalable Supervision, Shen Fan

Dissertations

Digital three-dimensional (3D) models are central to engineering design, analysis, and manufacturing, but learning pipelines for man-made geometry often operate on sampled carriers that do not preserve all of the structure present in exact CAD representations. This dissertation studies learning-based reconstruction and part decomposition for structured man-made 3D geometry, from general object benchmarks to CAD-derived datasets, with a focus on neural implicit representations trained from signed-distance samples, point clouds, and tessellated meshes. The goal is to make these models more accurate, more part-aware, and more consistently supervised.

First, signed distance function (SDF) reconstruction with implicit neural representations is improved through …

Parameter Density Estimation For Cardiac Electrophysiology Models Using Data Consistent Deep Learning, Michael Luo

Dissertations

Mathematical models of biological rhythms and excitable systems can provide insights into mechanisms underlying cardiac electrical dynamics. However, estimating the parameters of these models from experimental observations is often difficult due to noise, heterogeneity, and unobserved variables. For example, in an electrocardiogram (ECG) recording, information about the electrical properties of different regions of the heart is compressed into a single voltage trace. Additionally, variability within these signals may contain important information about population heterogeneity, regional differences in electrophysiology, and time-dependent modulation.

This dissertation develops, explores, and evaluates methods that perform feature-based distributional inference for complex nonlinear dynamical systems. The objective …

Robustness Of Ai-Driven Histopathology Under Real-World Adversarial Examples, Ruchik N. Yajnik

Theses

This robustness of histopathology classification models under adversarial and real-world perturbations resembling clinical artifacts is being investigated.

Using whole-slide images from the CAMELYON17 cohort, four representative architectures—ResNet-18, ResNet-50, HIPT-2MLP, and ViT-B/16 —are benchmarked across controlled pixel-level distortions and artifact-like transformations. Adversarial methods include iterative Fast Gradient Sign, Projected Gradient Descent, Salt-and-Pepper noise, and the Adversarial Watermark—Stain Shift (AWSS). Three defense strategies—Randomized Smoothing, Adversarial Training, and an Artifact Detector—are evaluated for their ability to preserve diagnostic accuracy and model reliability. Structured perturbations consistently degrade performance, with transformer-based models showing the greatest sensitivity. The benchmark developed here offers a reproducible framework for …

Predicting The Outcome Of Ischemic Hepatitis With Real-Patient Data Using Machine Learning Tools, Christiana Beard, Madison Utterback, Olcay Akman, Priya Kohli, William M. Lee, Aditi Ghosh

Spora: A Journal of Biomathematics

Ischemic hepatitis (IH) results from shock-related conditions that impair oxygenated blood flow to the liver, causing hepatocyte death. Diagnosis relies largely on clinical history due to the absence of specific diagnostic tests and limited ability to predict outcomes. This study applies machine learning methods to real-world IH patient data to improve outcome prediction. Biomedical indicators analyzed include creatinine, international normalized ratio (INR), aspartate aminotransferase (AST), alanine transaminase (ALT), and bilirubin. Data were collected from multiple U.S. centers through the Acute Liver Failure Study Group (ALFSG), a multicenter network focused on this rare condition. We implemented logistic regression, regression tree methods …

Decision Making At Triage Classification Using Svm With Smote Technique, Mehanas Shahul, Pushpalatha Kp

Northeast Journal of Complex Systems (NEJCS)

The efficient functioning of triage gates in overcrowded emergency departments (EDs) occurs in the context of the complex adaptive system (CAS) framework, where diverse system elements – patients, medical personnel, resources, patients’ inflow patterns, and patients themselves – simultaneously and dynamically influence the decision process. This study addresses the automated incorporation of machine learning triage algorithms as part of the system triage process to support automated classified risk-level recognition based on a limited set of vital signs. Patients are dynamically subsumed under high and low-risk categories enhanced by sensitivity, which enables optimal diagnosis and triage response to the critical clinician …

Optimizing Gated Rnns, Joshua Paul Fechete

Honors Projects

Gated recurrent neural networks such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU) help fix instability present in normal recurrent neural networks. This allows them to be used for various real-world tasks, and due to their architecture, they are uniquely qualified to handle variable sized input such as text. However, even before training can begin on a machine learning model, various hyperparameters must be chosen to decide how the model will be architectured. Choosing good hyperparameters is vital for creating a model that performs well but is not larger and more computationally expensive to run than it needs …

Applicability Of Machine Learning For Shear-Wave Velocity Prediction From Conventional Well Logs: The Lsboost Approach, Rahmat Catur Wibowo, Fadsyah Muhammad Arby, Bagus Sapto Mulyatno, Ordas Dewanto, Isti Nur Kumalasari, Muh Sarkowi

Turkish Journal of Earth Sciences

Shear-wave velocity (Vs) is one of the most critical parameters for determining geomechanical properties and basin overpressure. However, assessing Vs via techniques like core analysis requires considerable effort and expense. This study predicts Vs using several approaches and compares the accuracy levels of all models. For this objective, the multiple linear regression, multiple linear stepwise regression, support vector machine, and least-squares boost (LSBoost) methodologies were selected. The six well-logging data inputs of density (RHOB), gamma-ray (GR), deep resistivity (ILD), acoustic wave velocity (Vp), shale volume (VCL), and water saturation (SW) were selected as effective variables, whereas Vs was regarded as …

Accurate Diagnosis Of Diseases By A Novel Ai Pipeline Based On Feature Extraction, Feature Ranking, And Feature Selection From Medical Images, Tuğba Nur Bozkurt, Mehmet Emi̇n Yüksel

Turkish Journal of Electrical Engineering and Computer Sciences

The rapid growth of the global population has led to a substantial increase in the number of patients, while the availability of healthcare professionals has not expanded at a comparable rate. This imbalance highlights the urgent need for efficient and reliable computer-aided decision support systems that can reduce clinical workload while maintaining high diagnostic accuracy. In this study, a novel and systematically integrated artificial intelligence-based pipeline is proposed for medical image classification, combining statistical significance-driven feature ranking with evolutionary feature selection in a unified framework. The proposed pipeline consists of four sequential stages: feature extraction, ranking, selection, and classification. Features …

A Descriptive Analysis Of Plant Leaf Disease Detection Using Machine Learning And Deep Learning Models: A Systematic Review, Arzoo Chamoli, Anuj Kumar

Turkish Journal of Electrical Engineering and Computer Sciences

Plant leaf disease detection (PLDD) is a growing active research area with burgeoning practical applications across various sectors such as agricultural monitoring, food security, and environmental conservation. Accurate segmentation and classification of plant leaf diseases remains a key challenge in the field of plant leaf disease prediction. The challenge demands automated methods for the plant disease identification because it needs to develop better crop management systems, which will boost agricultural production. In this article, we provide a systematic review of various machine learning (ML) and deep learning (DL) methods extensively used for PLDD. The review strategy follows a formal protocol, …

Prescribing Company Action Through Machine Learning And Ai, Breck T. Husong

Data Science Undergraduate Honors Theses

The purpose of this research is to implement an OpenAI Reinforced Learning prescription-giving model for improving sales on a week-by-week basis. The data used comes from a segment of High Impact Analytics’s sales data that has been anonymized for proprietary reasons. The features among the data include inventory numbers, shipments in transit, total quantity and dollars of products sold each week for the past 2 years, all aggregated at the store-item-week level. In order to build this model, Tigramite, a causal discovery model combined with prediction models XGBoost, Linear Regression, Ridge Regression, Lasso Regression, Scikit-learn’s MLP, and Keras’s Neural Model …