Digital Commons Network^™

Leveraging The Inductive Bias Of Large Language Models For Abstract Textual Reasoning, Christopher Michael Rytting

Theses and Dissertations

Large natural language models (such as GPT-2 or T5) demonstrate impressive abilities across a range of general NLP tasks. Here, we show that the knowledge embedded in such models provides a useful inductive bias, not just on traditional NLP tasks, but also in the nontraditional task of training a symbolic reasoning engine. We observe that these engines learn quickly and generalize in a natural way that reflects human intuition. For example, training such a system to model block-stacking might naturally generalize to stacking other types of objects because of structure in the real world that has been partially captured by …

Traffic Time Headway Prediction And Analysis: A Deep Learning Approach, Saumik Sakib Bin Masud

Theses and Dissertations

In the modern world of Intelligent Transportation System (ITS), time headway is a key traffic flow parameter affecting ITS operations and planning. Defined as “the time difference between any two successive vehicles when they cross a given point”, time headway is used in various traffic and transportation engineering research domains, such as capacity analysis, safety studies, car-following, and lane-changing behavior modeling, and level of service evaluation describing stochastic features of traffic flow. Advanced travel and headway information can also help road users avoid traffic congestion through dynamic route planning, for instance. Hence, it is crucial to accurately model headway distribution …

A Targeted Adversarial Attack On Support Vector Machine Using The Boundary Line, Yessenia Rodriguez

Theses and Dissertations

In this thesis, a targeted adversarial attack is explored on a Support Vector Machine (SVM). SVM is defined by creating a separating boundary between two classes. Using a target class, any input can be modified to cross the “boundary line,” making the model predict the target class. To limit the modification, a percentage of an image of the target class is used to get several random sections. Using these sections, the input will be moved in small steps closer to the boundary point. The section that took the least number of steps to cause the model to predict the target …

Semantic-Driven Unsupervised Image-To-Image Translation For Distinct Image Domains, Wesley Ackerman

Theses and Dissertations

We expand the scope of image-to-image translation to include more distinct image domains, where the image sets have analogous structures, but may not share object types between them. Semantic-Driven Unsupervised Image-to-Image Translation for Distinct Image Domains (SUNIT) is built to more successfully translate images in this setting, where content from one domain is not found in the other. Our method trains an image translation model by learning encodings for semantic segmentations of images. These segmentations are translated between image domains to learn meaningful mappings between the structures in the two domains. The translated segmentations are then used as the basis …

Joint 1d And 2d Neural Networks For Automatic Modulation Recognition, Luis M. Rosario Morel

Theses and Dissertations

The digital communication and radar community has recently manifested more interest in using data-driven approaches for tasks such as modulation recognition, channel estimation and distortion correction. In this research we seek to apply an object detector for parameter estimation to perform waveform separation in the time and frequency domain prior to classification. This enables the full automation of detecting and classifying simultaneously occurring waveforms. We leverage a lD ResNet implemented by O'Shea et al. in [1] and the YOLO v3 object detector designed by Redmon et al. in [2]. We conducted an in depth study of the performance of these …

Deep Learning To Predict Ocean Seabed Type And Source Parameters, David Franklin Van Komen

Theses and Dissertations

In the ocean, light from the surface dissipates quickly leaving sound the only way to see at a distance. Different sediment types on the ocean floor and water properties like salinity, temperature, and ocean depth all change how sound travels across long distances. Hard sediment types, such as sand and bedrock, are highly reflective while softer sediment types, such as mud, are more absorptive and change the received sound upon arrival. Unfortunately, the vast majority of the ocean floor is not mapped and the expenses involved in creating such a map are far too great. Traditional signal processing methods in …

Advanced In-Situ Layer-Wise Quality Control For Laser-Based Additive Manufacturing Using Image Sequence Analysis, Mehrnaz Noroozi Esfahani

Theses and Dissertations

Quality assurance has been one of the major challenges in laser-based additive manufacturing (AM) processes. This study proposes a novel process modeling methodology for layer-wise in-situ quality monitoring based on image series analysis. An image-based autoregressive (AR) model has been proposed based on the image registration function between consecutively observed thermal images. Image registration is used to extract melt pool location and orientation change between consecutive images, which contains sensing stability information. Subsequently, a Gaussian process model is used to characterize the spatial correlation within the error matrix. Finally, the extracted features from the aforementioned processes are jointly used for …

Towards Generation Of Creative Software Requirements, Quoc Anh Do Jr

Theses and Dissertations

Increasingly competitive software industry, where multiple systems serve the same application domain and compete for customers, favors software with creative features. To promote software creativity, research has proposed multi-day workshops with experienced facilitators, and semi-automated tools to provide a limited support for creative thinking. Such approach is either time consuming and demands substantial involvement from analysts with creative abilities, or useful only for existing large-scale software with a rich issue tracking system. In this dissertation, we present different approaches leveraging advanced natural language processing and machine learning techniques to provide automated support for generating creative software requirements with minimal human …

An Exploration Of Success Factors In The Healthcare Supply Chain, Matthew Tidwell

Theses and Dissertations

This research builds off previous research conducted in 2009 which included a survey of healthcare professionals assessing their organization’s levels of supply chain maturity (SCM) and data standard readiness (DSR) from 1 to 5 [Smith, 2011]. With the survey data, Smith developed a 0-1 quadratic program to conserve the maximum amount of survey data while removing non-responses. This research uses the quadratic program as well as other machine learning algorithms and analysis methods to investigate what factors contribute to an organization’s SCM and DSR levels the most. No specific factors were found; however, different levels of prediction accuracy were achieved …

Identifying And Mitigating Heat Stress Of Grazing Dairy Cattle Using Shade And Sprinklers, Carly Becker

Theses and Dissertations

Animal welfare, reproduction, and milk production can be negatively affected when dairy cattle experience heat stress. Dairy cows in southern latitudes spend nearly 4 to 6 months in a state of heat stress. Animal health professionals and dairy producers use changes in physiological responses and behavioral patterns of cows as a tool for identifying poor health and welfare in periods of heat stress. The objectives of this study were to monitor the effects of heat stress on grazing dairy cows provided with shade or sprinklers by comparing various physiological indices of heat stress, and to, design and utilize a heat …

Detection Of Stealthy False Data Injection Attacks Against State Estimation In Electric Power Grids Using Deep Learning Techniques, Qingyu Ge

Theses and Dissertations

Since communication technologies are being integrated into smart grid, its vulnerability to false data injection is increasing. State estimation is a critical component which is used for monitoring the operation of power grid. However, a tailored attack could circumvent bad data detection of the state estimation, thus disturb the stability of the grid. Such attacks are called stealthy false data injection attacks (FDIAs). This thesis proposed a prediction-based detector using deep learning techniques to detect injected measurements. The proposed detector adopts both Convolutional Neural Networks and Recurrent Neural Networks, making full use of the spatial-temporal correlations in the measurement data. …

Information Retrieval Of Opioid Dependence Medications Reviews From Health-Related Social Media, Seyedeh Samaneh Omranian

Theses and Dissertations

Social media provides a convenient platform for patients to share their drug usage experience with others; consequently, health researchers can leverage this potential data to gain valuable information about users’ drug satisfaction. Since the 1990s, opioid drug abuse has become a national crisis. In order to reduce the dependency of opioids, several drugs have been presented to the market, but little is known about patient satisfaction with these treatments. Sentiment analysis is a method to measure and interpret patients’ satisfaction. In the first phase of this study, we aimed to utilize social media posts to predict patients’ sentiment towards opioid …

Signal Processing Combined With Machine Learning For Biomedical Applications, Md Shakhawat Hossain

Theses and Dissertations

The Master’s thesis is comprised of four projects in the realm of machine learning and signal processing. The abstract of the thesis is divided into four parts and presented as follows,

Abstract 1: A Kullback-Leibler Divergence-Based Predictor for Inter-Subject Associative BCI.

Inherent inter-subject variability in sensorimotor brain dynamics hinders the transferability of brain-computer interface (BCI) model parameters across subjects. An individual training session is essential for effective BCI control to compensate for variability. We report a Kullback-Leibler Divergence (KLD)-based predictor for inter-subject associative BCI. An online dataset comprising left/right hand, both feet, and tongue motor imagery tasks was used to …

Age-Suitability Prediction For Literature Using Deep Neural Networks, Eric Robert Brewer

Theses and Dissertations

Digital media holds a strong presence in society today. Providers of digital media may choose to obtain a content rating for a given media item by submitting that item to a content rating authority. That authority will then issue a content rating that denotes to which age groups that media item is appropriate. Content rating authorities serve publishers in many countries for different forms of media such as television, music, video games, and mobile applications. Content ratings allow consumers to quickly determine whether or not a given media item is suitable to their age or preference. Literature, on the other …

Theoretical Investigation Of The Biomass Conversion On Transition Metal Surfaces Based On Density Functional Theory Calculations And Machine Learning, Wenqiang Yang

Theses and Dissertations

During the past decades, heterogenous catalyzed conversion of biomass to hydrocarbons with similar or identical properties to conventional fossil fuels has gained significantly academic and industrial interest. However, the conventional heterogeneous catalysts such as sulfided NiMo/Al₂O₃ and CoMo/Al₂O₃ used have various drawbacks, such as short catalyst lifetime and high sulfur content of product. To overcome the limitations of the conventional sulfided catalysts, new catalysts must be developed, which requires a better understanding of the reaction mechanism of the biomass conversion. Based on density functional theory, in this thesis, we reported a computational calculation study …

Evaluation Of Temporal Damage Progression In Concrete Structures Affected By Asr Using Data-Driven Methods, Vafa Soltangharaei

Theses and Dissertations

Alkali-silica reaction (ASR) is a chemical reaction, which causes damage in concrete structures such as bridges, dams, and nuclear containments and powerplant structures. The ASR-induced damage may endanger the integrity and serviceability of structures. Several methods such as visual inspection, petrographic analysis, demountable mechanical strain gauges, and cracking index have been utilized for study the effect of ASR on structures, which are not always efficient in early damage detection and some are destructive and prohibited in nuclear structures. Nondestructive methods and structural health monitoring techniques can be alternatives for the condition assessment of structures. Among the nondestructive methods, acoustic emission …

Towards Cooperating In Repeated Interactions Without Repeating Structure, Huy Pham

Theses and Dissertations

A big challenge in artificial intelligence (AI) is creating autonomous agents that can interact well with other agents over extended periods of time. Most previously developed algorithms have been designed in the context of Repeated Games, environments in which the agents interact in the same scenario repeatedly. However, in most real-world interactions, relationships between people and autonomous agents consist of sequences of distinct encounters with different incentives and payoff structures. Therefore, in this thesis, we consider Interaction Games, which model interactions in which the scenario changes from encounter to encounter, often in ways that are unanticipated by the players. For …

Optimized 3d Reconstruction For Infrastructure Inspection With Automated Structure From Motion And Machine Learning Methods, Samuel Arce Munoz

Theses and Dissertations

Infrastructure monitoring is being transformed by the advancements on remote sensing, unmanned vehicles and information technology. The wide interaction among these fields and the availability of reliable commercial technology are helping pioneer intelligent inspection methods based on digital 3D models. Commercially available Unmanned Aerial Vehicles (UAVs) have been used to create 3D photogrammetric models of industrial equipment. However, the level of automation of these missions remains low. Limited flight time, wireless transfer of large files and the lack of algorithms to guide a UAV through unknown environments are some of the factors that constraint fully automated UAV inspections. This work …

A Reinforcement Learning Approach To Sequential Acceptance Sampling As A Critical Success Factor For Lean Six Sigma, Hani A. Khalil

Theses and Dissertations

In the 21st century, globalization coupled with technological advancement and free trade has created competition among various businesses enterprises. This competition has led many businesses to adopt various management techniques such as acceptance sampling aimed at transforming their processes in order to remain competitive in the global market and adapt to new market demands. The successful implementation of acceptance sampling is highly dependent on what the academic literature refers to as acceptance sampling optimization. A literature review on the optimization of acceptance sampling has not shown any work that studied whether acceptance sampling and machine learning (ML) plans can be …

Old Dogs, New Tricks: Authoritarian Regime Persistence Through Learning, Nicholas Ryan Davis

Theses and Dissertations

How does diffusion lead to authoritarian regime persistence? Political decisions, regardless of what the actors involved might believe or espouse, do not happen in isolation. Policy changes, institutional alterations, regime transitions-- these political phenomena are all in some part a product of diffusion processes as much as they are derived from internal determinants. As such, political regimes do not exist in a vacuum, nor do they ignore the outside world. When making decisions about policy and practice, we should expect competent political actors to take a look at the wider external world. This dissertation project presents a theory of regime …

Determinants Of Safety Outcomes In Organizations: Exploring O*Net Data To Predict Occupational Accident Rates, Lavanya Shravan Kumar

Theses and Dissertations

Workplace safety is of utmost importance given the regular occurrence of both fatal and nonfatal occupational injuries all around the world. Although research in this area is hugely prevalent, it is focused mainly on safety climate and lacks an integrated approach when examining predictors of safety outcomes. The development of an occupational risk factor that predicts safety outcomes will aid in understanding the relative importance of different factors that contribute to safety and help organizations target their safety programs and interventions efficiently. The present study is an exploratory analysis utilizing publicly available O*NET data (work activities, work context features, and …

Airfoil Analysis And Design Using Surrogate Models, Nicholas Alexander Michael

Theses and Dissertations

A study was performed to compare two different methods for generating surrogate models for the analysis and design of airfoils. Initial research was performed to compare the accuracy of surrogate models for predicting the lift and drag of an airfoil with data collected from highidelity simulations using a modern CFD code along with lower-order models using a panel code. This was followed by an evaluation of the Class Shape Trans- formation (CST) method for parameterizing airfoil geometries as a prelude to the use of surrogate models for airfoil design optimization and the implementation of software to use CST to modify …

Mechanical And Thermal Behavior Of Multiscale Bi-Nano-Composites Using Experiments And Machine Learning Predictions, Vahid Daghigh

Theses and Dissertations

The mechanical and thermal properties of natural short latania fiber (SLF)-reinforced poly(propylene)/ethylene-propylene-diene-monomer (SLF/PP/EPDM) bio-composites reinforced with nano-clays (NCs), pistachio shell powders (PSPs), and/or date seed particles (DSPs) were studied using experiments and machine learning (ML) predictions. This dissertation embraces three related investigations: (1) an assessment of maleated polypropylene (MAPP) coupling agent on mechanical and thermal behavior of SLF/PP/EPDM composites, (2) heat deflection temperature (HDT) of bio-nano-composites using experiments and ML predictions, and (3) fracture toughness ML predictions of short fiber, nano- and micro-particle reinforced composites. The first project (Chapter 2) investigates the influence of MAPP on tensile, bending, Charpy impact …

A Machine Learning Framework For Prediction Of Diagnostic Trouble Codes In Automobiles, Mohan Kopuru

Theses and Dissertations

Predictive Maintenance is an important solution to the rising maintenance costs in the industries. With the advent of intelligent computer and availability of data, predictive maintenance is seen as a solution to predict and prevent the occurrence of the faults in the different types of machines. This thesis provides a detailed methodology to predict the occurrence of critical Diagnostic Trouble codes that are observed in a vehicle in order to take necessary maintenance actions before occurrence of the fault in automobiles using Convolutional Neural Network architecture.

Glacier Change Assessment Of The Columbia Icefield In The Canadian Rocky Mountains, Canada (1985 – 2018), Adjoa Dwamena Intsiful

Theses and Dissertations

Glaciers adjust their sizes as a response to changing climatic conditions which make them a good indicator of climate change. Remote-sensing based glacier monitoring provides a robust way to inventory the health of glaciers and are estimated as a measure of changes in their area, length, volume and mass balance over a period. This research uses remote sensing methods to map glacier extents from satellite images and explores the efficacy of three machine learning algorithms for accurate glacier classification. The results indicated that the Columbia icefield lost 42 km2 of its area cover between 1985 and 2018. It was observed …

Automated Digit Recognition On Sound Pressure Level Meters Based On Deep Learning, Che-Wei Tung

Theses and Dissertations

Sound pressure level (SPL) meter is one of the useful devices used for measuring the sound level pressure. The measurement device displays the SPL value in decibels (dB) on a standard LCD screen (no backlight). We could base on the digit number shown on the LCD screen to do some adjustments or evaluations. Thus, SPL has been widely used in several fields to quantify different noise, such as industrial, environmental, and aircraft noise. However, in my basic knowledge, there is no previous study used machine learning to auto-recognize the digit on the SPL meter. This thesis presents a novel system …

Chaotic Model Prediction With Machine Learning, Yajing Zhao

Theses and Dissertations

Chaos theory is a branch of modern mathematics concerning the non-linear dynamic systems that are highly sensitive to their initial states. It has extensive real-world applications, such as weather forecasting and stock market prediction. The Lorenz system, defined by three ordinary differential equations (ODEs), is one of the simplest and most popular chaotic models. Historically research has focused on understanding the Lorenz system's mathematical characteristics and dynamical evolution including the inherent chaotic features it possesses. In this thesis, we take a data-driven approach and propose the task of predicting future states of the chaotic system from limited observations. We explore …

Lightning Prediction For Space Launch Using Machine Learning Based Off Of Electric Field Mills And Lightning Detection And Ranging Data, Anson Cheng

Theses and Dissertations

Kennedy Space Center and Cape Canaveral Air Station, FL, where the Air Force conducts space launches, are in an area of frequent lightning strikes, which is main obstacle in meeting launch goals. The 45th Weather Squadron (45th WS) ensures that any weather safety requirements are met during pre-launch and actual space launch. Using only summer months from three years’ worth of lightning detection and ranging (LDAR) and electric field mill (EFM) data from KSC, several feedforward neural networks are constructed. Separate models are built for each EFM and trained by adjusting parameters to forecast lightning 30 minutes out in the …

Event-Based Visual-Inertial Odometry Using Smart Features, Zachary P. Friedel

Theses and Dissertations

Event-based cameras are a novel type of visual sensor that operate under a unique paradigm, providing asynchronous data on the log-level changes in light intensity for individual pixels. This hardware-level approach to change detection allows these cameras to achieve ultra-wide dynamic range and high temporal resolution. Furthermore, the advent of convolutional neural networks (CNNs) has led to state-of-the-art navigation solutions that now rival or even surpass human engineered algorithms. The advantages offered by event cameras and CNNs make them excellent tools for visual odometry (VO). This document presents the implementation of a CNN trained to detect and describe features within …

Retiming Smoke Simulation Using Machine Learning, Samuel Charles Gérard Giraud Carrier

Theses and Dissertations

Art-directability is a crucial aspect of creating aesthetically pleasing visual effects that help tell stories. A particularly common method of art direction is the retiming of a simulation. Unfortunately, the means of retiming an existing simulation sequence which preserves the desired shapes is an ill-defined problem. Naively interpolating values between frames leads to visual artifacts such as choppy frames or jittering intensities. Due to the difficulty in formulating a proper interpolation method we elect to use a machine learning approach to approximate this function. Our model is based on the ODE-net structure and reproduces a set of desired time samples …