Computer Sciences Commons^™

Imagining New Futures Beyond Predictive Systems In Child Welfare: A Qualitative Study With Impacted Stakeholders, Logan Stapleton, Min Hun Lee, Diana Qing, Marya Wright, Alexandra Chouldechova, Ken Holstein, Zhiwei Steven Wu, Haiyi Zhu

Research Collection School Of Computing and Information Systems

Child welfare agencies across the United States are turning to datadriven predictive technologies (commonly called predictive analytics) which use government administrative data to assist workers’ decision-making. While some prior work has explored impacted stakeholders’ concerns with current uses of data-driven predictive risk models (PRMs), less work has asked stakeholders whether such tools ought to be used in the first place. In this work, we conducted a set of seven design workshops with 35 stakeholders who have been impacted by the child welfare system or who work in it to understand their beliefs and concerns around PRMs, and to engage them …

Local Learning Algorithms For Stochastic Spiking Neural Networks, Bleema Rosenfeld

Dissertations

This dissertation focuses on the development of machine learning algorithms for spiking neural networks, with an emphasis on local three-factor learning rules that are in keeping with the constraints imposed by current neuromorphic hardware. Spiking neural networks (SNNs) are an alternative to artificial neural networks (ANNs) that follow a similar graphical structure but use a processing paradigm more closely modeled after the biological brain in an effort to harness its low power processing capability. SNNs use an event based processing scheme which leads to significant power savings when implemented in dedicated neuromorphic hardware such as Intel’s Loihi chip.

This work …

Adversarially Robust And Accurate Machine Learning For Image Classification, Yanan Yang

Dissertations

Machine learning techniques in medical imaging systems are accurate, but minor perturbations in the data known as adversarial attacks can fool them. These attacks make the systems vulnerable to fraud and deception, and thus a significant challenge has been posed in practice. This dissertation presents the gradient-free trained sign activation networks to detect and deter adversarial attacks on medical imaging AI (Artificial Intelligence) systems. Experimental results show a higher distortion value is required to attack the proposed model than other state-of-the-art models on brain MRI (Magnetic resonance imaging), Chest X-ray, and histopathology image datasets. Moreover, the proposed models outperform the …

Un-Fair Trojan: Targeted Backdoor Attacks Against Model Fairness, Nicholas Furth

Theses

Machine learning models have been shown to be vulnerable against various backdoor and data poisoning attacks that adversely affect model behavior. Additionally, these attacks have been shown to make unfair predictions with respect to certain protected features. In federated learning, multiple local models contribute to a single global model communicating only using local gradients, the issue of attacks become more prevalent and complex. Previously published works revolve around solving these issues both individually and jointly. However, there has been little study on the effects of attacks against model fairness. Demonstrated in this work, a flexible attack, which we call Un-Fair …

Deep-Learning-Incorporated Augmented Reality Application For Engineering Lab Training, John Estrada, Sidike Paheding, Xiaoli Yang, Quamar Niyaz

Michigan Tech Publications, Part 1

Deep learning (DL) algorithms have achieved significantly high performance in object detection tasks. At the same time, augmented reality (AR) techniques are transforming the ways that we work and connect with people. With the increasing popularity of online and hybrid learning, we propose a new framework for improving students’ learning experiences with electrical engineering lab equipment by incorporating the abovementioned technologies. The DL powered automatic object detection component integrated into the AR application is designed to recognize equipment such as multimeter, oscilloscope, wave generator, and power supply. A deep neural network model, namely MobileNet-SSD v2, is implemented for equipment detection …

A Systematic Review On Machine Learning Models For Online Learning And Examination Systems, Sanaa Kaddoura, Daniela Elena Popescu, Jude D. Hemanth

All Works

Examinations or assessments play a vital role in every student’s life; they determine their future and career paths. The COVID pandemic has left adverse impacts in all areas, including the academic field. The regularized classroom learning and face-to-face real-time examinations were not feasible to avoid widespread infection and ensure safety. During these desperate times, technological advancements stepped in to aid students in continuing their education without any academic breaks. Machine learning is a key to this digital transformation of schools or colleges from real-time to online mode. Online learning and examination during lockdown were made possible by Machine learning methods. …

Real Time Call-Flagging System To Respond To Suicidal Ideation In Call Centers, Vishnu Menon, Joseph Carrigan, Charles Floeder, Thomas Walton, Devin Mcguire

Honors Program: Senior Projects (Public)

The 2021-2022 Signature Performance Design Studio team developed a live audio call-flagging system that enables faster responses and new response pathways to veteran crises by call service representatives and their management team. Using a custom made deep learning model, live audio streaming server, and Teams broadcasting add-on, the system empowers Signature Performance call service representatives to make quicker and more well informed decisions to provide veteran’s the best care possible.

The Contribution Of Ethical Governance Of Artificial Intelligence & Machine Learning In Healthcare, Tina Nguyen

Electronic Theses and Dissertations

With the Internet Age and technology progressively advancing every year, the usage of Artificial Intelligence (AI) along with Machine Learning (ML) algorithms has only increased since its introduction to society. Specifically, in the healthcare field, AI/ML has proven to its end-users how beneficial its assistance has been. However, despite its effectiveness and efficiencies, AI/ML has also been under scrutiny due to its unethical outcomes. As a result of this, two polarizing views are typically debated when discussing AI/ML. One side believes that AI/ML usage should continue regardless of its unsureness, while the other side argues that this technology is too …

Computational Approaches To Facilitate Automated Interchange Between Music And Art, Rao Hamza Ali

Computational and Data Sciences (PhD) Dissertations

Recently, there has been a tremendous increase in generating and synthesizing music and art using various computational techniques. An area that is still under-researched, however, is how one medium can be converted into the other, while maintaining the overall aesthetics. Over the last few centuries, artists, composers, and scholars, have attempted to use substitute one form of art for the other: by proposing techniques where music notes are synonymous to colors, by inventing instruments that combine the aesthetics of music and visual art, and by incorporating the two media in live performances. A widely accepted computational approach, for the conversion, …

Intraday Algorithmic Trading Using Momentum And Long Short-Term Memory Network Strategies, Andrew R. Whitinger Ii

Undergraduate Honors Theses

Intraday stock trading is an infamously difficult and risky strategy. Momentum and reversal strategies and long short-term memory (LSTM) neural networks have been shown to be effective for selecting stocks to buy and sell over time periods of multiple days. To explore whether these strategies can be effective for intraday trading, their implementations were simulated using intraday price data for stocks in the S&P 500 index, collected at 1-second intervals between February 11, 2021 and March 9, 2021 inclusive. The study tested 160 variations of momentum and reversal strategies for profitability in long, short, and market-neutral portfolios, totaling 480 portfolios. …

Machine Learning Methods For Statistical Analysis And Representation Learning On Neuroimaging Data, Fan Yang

Computer Science and Engineering Dissertations - Archive

With the recent advance and widespread adoption of imaging technological innovations, clinical practitioner and scientists can easily acquire and store a large amount of various neuroimaging modalities, such as Diffusion Tensor Imaging (DTI), Magnetic Resonance Imaging (MRI), resting-state functional MRI (rs-fMRI) and Positron Emission Tomography (PET), etc. These novel imaging data sources cover a rich amount of factors that influence patients' cognitive health, offer an objective view of patients at unprecedented multi-resolution for the understanding of brain structure and function, and have the significant potential to improve healthcare by aiding better decision-making in diagnosing, monitoring and treating diseases. Machine Learning …

Deep Learning Object-Based Detection Of Manufacturing Defects In X-Ray Inspection Imaging, Juan C. Parducci

Mechanical & Aerospace Engineering Theses & Dissertations

Current analysis of manufacturing defects in the production of rims and tires via x-ray inspection at an industry partner’s manufacturing plant requires that a quality control specialist visually inspect radiographic images for defects of varying sizes. For each sample, twelve radiographs are taken within 35 seconds. Some defects are very small in size and difficult to see (e.g., pinholes) whereas others are large and easily identifiable. Implementing this quality control practice across all products in its human-effort driven state is not feasible given the time constraint present for analysis.

This study aims to identify and develop an object detector capable …

Machine Learning Classification Of Digitally Modulated Signals, James A. Latshaw

Electrical & Computer Engineering Theses & Dissertations

Automatic classification of digitally modulated signals is a challenging problem that has traditionally been approached using signal processing tools such as log-likelihood algorithms for signal classification or cyclostationary signal analysis. These approaches are computationally intensive and cumbersome in general, and in recent years alternative approaches that use machine learning have been presented in the literature for automatic classification of digitally modulated signals. This thesis studies deep learning approaches for classifying digitally modulated signals that use deep artificial neural networks in conjunction with the canonical representation of digitally modulated signals in terms of in-phase and quadrature components. Specifically, capsule networks are …

Data-Driven Framework For Understanding & Modeling Ride-Sourcing Transportation Systems, Bishoy Kelleny

Civil & Environmental Engineering Theses & Dissertations

Ride-sourcing transportation services offered by transportation network companies (TNCs) like Uber and Lyft are disrupting the transportation landscape. The growing demand on these services, along with their potential short and long-term impacts on the environment, society, and infrastructure emphasize the need to further understand the ride-sourcing system. There were no sufficient data to fully understand the system and integrate it within regional multimodal transportation frameworks. This can be attributed to commercial and competition reasons, given the technology-enabled and innovative nature of the system. Recently, in 2019, the City of Chicago the released an extensive and complete ride-sourcing trip-level data for …

Identifying Functional And Non-Functional Software Requirements From User App Reviews And Requirements Artifacts, Dev Jayant Dave

Theses, Dissertations and Culminating Projects

This thesis proposes and evaluates Machine Learning (ML) based data models to identify and isolate software requirements from datasets containing user app review statements. The ML models classify user app review statements into Functional Requirements (FRs), Non-Functional Requirements (NFRs), and Non-Requirements (NRs). This proposed approach consisted of creating a novel hybrid dataset that contains software requirements from Software Requirements Specification (SRS) documents and user app reviews. The Support Vector Machine (SVM), Stochastic Gradient Descent (SGD), and Random Forest (RF) ML algorithms combined with the term frequency-inverse document frequency (TF-IDF) natural language processing (NLP) technique were implemented on the hybrid dataset. …

Exposing Gan-Generated Faces Using Deep Neural Network, Hui Guo

Legacy Theses & Dissertations (2009 - 2024)

Generative adversarial network (GAN) generated high-realistic human faces are visually challenging to discern from real ones. They have been used as profile images for fake social media accounts, which leads to high negative social impacts.In this work, we explore a universal physiological cue of the eye, namely the pupil shape consistency, to identify GAN-generated faces reliably. We show that GAN-generated faces can be exposed via irregular pupil shapes. This phenomenon is caused by the lack of physiological constraints in the GAN models. We demonstrate that such artifacts exist widely in high-quality GAN-generated faces. We design an automatic method to segment …

The Quest For New Music: A Recommendation Algorithm For Spotify Users, Ian Curtis

Honors Projects

Music is one of the rare forms of communication that can be understood on a profound level by anyone; it has the power to cause significant emotional effects, to spark inspiration, to ignite change, to spread knowledge, and more, even regardless of song language. A popular subject of research in music pertains to recommendations; determining a song a listener would enjoy is not an easy task. Moreover, certain factors may influence a user's satisfaction with recommended songs and their likelihood to continue using a service. Focusing on the major streaming service Spotify, we build a K-Means clustering algorithm to recommend …

Computational Approaches To Understanding Subduction Zone Geodynamics, Surface Heat Flow, And The Metamorphic Rock Record, Buchanan C. Kerswell

Boise State University Theses and Dissertations

Pressure-temperature (PT) estimates from exhumed high-pressure (HP) metamorphic rocks and global surface heat flow observations evidently encode information about subduction zone thermal structure and the nature of mechanical and chemical processing of subducted materials along the interface between converging plates. Previous work demonstrates the possibility of decoding such geodynamic information by comparing numerical geodynamic models with empirical observations of surface heat flow and the metamorphic rock record. However, ambiguous interpretations can arise from this line of inquiry with respect to thermal gradients, plate coupling, and detachment and recovery of subducted materials. This dissertation applies a variety of computational techniques to …

New Debiasing Strategies In Collaborative Filtering Recommender Systems: Modeling User Conformity, Multiple Biases, And Causality., Mariem Boujelbene

Electronic Theses and Dissertations

Recommender Systems are widely used to personalize the user experience in a diverse set of online applications ranging from e-commerce and education to social media and online entertainment. These State of the Art AI systems can suffer from several biases that may occur at different stages of the recommendation life-cycle. For instance, using biased data to train recommendation models may lead to several issues, such as the discrepancy between online and offline evaluation, decreasing the recommendation performance, and hurting the user experience. Bias can occur during the data collection stage where the data inherits the user-item interaction biases, such as …

Beyond Accuracy In Machine Learning., Aneseh Alvanpour

Electronic Theses and Dissertations

Machine Learning (ML) algorithms are widely used in our daily lives. The need to increase the accuracy of ML models has led to building increasingly powerful and complex algorithms known as black-box models which do not provide any explanations about the reasons behind their output. On the other hand, there are white-box ML models which are inherently interpretable while having lower accuracy compared to black-box models. To have a productive and practical algorithmic decision system, precise predictions may not be sufficient. The system may need to have transparency and be able to provide explanations, especially in applications with safety-critical contexts …

Modeling And Debiasing Feedback Loops In Collaborative Filtering Recommender Systems., Sami Khenissi

Electronic Theses and Dissertations

Artificial Intelligence (AI)-driven recommender systems have been gaining increasing ubiquity and influence in our daily lives, especially during time spent online on the World Wide Web or smart devices. The influence of recommender systems on who and what we can find and discover, our choices, and our behavior, has thus never been more concrete. AI can now predict and anticipate, with varying degrees of accuracy, the news article we will read, the music we will listen to, the movies we will watch, the transactions we will make, the restaurants we will eat in, the online courses we will be interested …

Hardware Isolation Approach To Securely Use Untrusted Gpus In Cloud Environments For Machine Learning, Lucas D. Hall

Theses and Dissertations

Machine Learning (ML) is now a primary method for getting useful information out of the immense volumes of data being generated and stored in society today. Useful data is a commodity for training ML models and those that need data for training are often not the owners of the data leading to a desire to use cloud-based services. Deep learning algorithms are best suited to run on a graphical processing unit (GPU) which presents a specific problem since the GPU is not a secure or trusted piece of hardware in the cloud computing environment.

In this paper, we will analyze …

Characterization And Mitigation Of False Information On The Web, Anu Shrestha

Boise State University Theses and Dissertations

Social media and Web sources have made information available, accessible, and shareable any time and anywhere nearly without friction. This information can be truthful, falsified, or can only be the opinion of the writer as users in such platforms are both information creators and consumers. In any case, it has the power to affect the decision of an individual, the beliefs of the society, activities, and the economy of the whole country. Thus, it is imperative to identify false information and mitigate the effects of false information that are ubiquitous across the Web and social media. Therefore, the main goal …

A Comprehensive Survey For Non-Intrusive Load Monitoring, Efe İsa Tezde, Eray Yildiz

Turkish Journal of Electrical Engineering and Computer Sciences

Energy-saving and efficiency are as important as benefiting from new energy sources to supply increasing energy demand globally. Energy demand and resources for energy saving should be managed effectively. Therefore, electrical loads need to be monitored and controlled. Demand-side energy management plays a vital role in achieving this objective. Energy management systems schedule an optimal operation program for these loads by obtaining more accurate and precise residential and commercial loads information. Different intellegent measurement applications and machine learning algorithms have been proposed for the measurement and control of electrical devices/loads used in buildings. Of these, nonintrusive load monitoring (NILM) is …

Automated Identification Of Libraries From Vulnerability Data: Can We Do Better?, Stefanus A. Haryono, Hong Jin Kang, Abhishek Sharma, Asankhaya Sharma, Andrew E. Santosa, Ming Yi Ang, David Lo

Research Collection School Of Computing and Information Systems

Software engineers depend heavily on software libraries and have to update their dependencies once vulnerabilities are found in them. Software Composition Analysis (SCA) helps developers identify vulnerable libraries used by an application. A key challenge is the identification of libraries related to a given reported vulnerability in the National Vulnerability Database (NVD), which may not explicitly indicate the affected libraries. Recently, researchers have tried to address the problem of identifying the libraries from an NVD report by treating it as an extreme multi-label learning (XML) problem, characterized by its large number of possible labels and severe data sparsity. As input, …

Dataset Evaluation For Data Trading Using Expected Loss And Homomorphic Encryption, Minsung Joo

Senior Honors Papers / Undergraduate Theses

Supervised machine learning suffers from the ``garbage-in garbage-out" phenomenon where the performance of a model is limited by the quality of the data. While a myriad of data is collected every second, there is no general rigorous method of evaluating the quality of a given dataset. This hinders fair pricing of data in scenarios where a buyer may look to buy data for use with machine learning. In this work, I propose using the expected loss corresponding to a dataset as a measure of its quality, relying on Bayesian methods for uncertainty quantification. Furthermore, I present a secure multi-party computation …

Development Of A Machine Learning-Based Financial Risk Control System, Zhigang Hu

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

With the gradual end of the COVID-19 outbreak and the gradual recovery of the economy, more and more individuals and businesses are in need of loans. This demand brings business opportunities to various financial institutions, but also brings new risks. The traditional loan application review is mostly manual and relies on the business experience of the auditor, which has the disadvantages of not being able to process large quantities and being inefficient. Since the traditional audit processing method is no longer suitable some other method of reducing the rate of non-performing loans and detecting fraud in applications is urgently needed …

Parallel Construction Of Random Forest On Gpu, Kennedy Senagi, Nicolas Jouandeau

All Peer-Reviewed Publications

There is tremendous growth of data generated from different industries, i.e., health, agriculture, engineering, etc. Consequently, there is demand for more processing power. Compared to computer processing units, general-purpose graphics processing units (GPUs) are rapidly emerging as a promising solution to achieving high performance and energy efficiency in various computing domains. Multiple forms of parallelism and complexity in memory access have posed a challenge in developing Random Forest (RF) GPU-based algorithm. RF is a popular and robust machine learning algorithm. In this paper, coarse-grained and dynamic parallelism approaches on GPU are integrated into RF(dpRFGPU). Experiment results of dpRFGPU are compared …

Structure-Aware Visualization Retrieval, Haotian Li, Yong Wang, Aoyu Wu, Huan Wei, Huamin. Qu

Research Collection School Of Computing and Information Systems

With the wide usage of data visualizations, a huge number of Scalable Vector Graphic (SVG)-based visualizations have been created and shared online. Accordingly, there has been an increasing interest in exploring how to retrieve perceptually similar visualizations from a large corpus, since it can benefit various downstream applications such as visualization recommendation. Existing methods mainly focus on the visual appearance of visualizations by regarding them as bitmap images. However, the structural information intrinsically existing in SVG-based visualizations is ignored. Such structural information can delineate the spatial and hierarchical relationship among visual elements, and characterize visualizations thoroughly from a new perspective. …

Data-Driven Models For Remaining Useful Life Estimation Of Aircraft Engines And Hard Disk Drives, Austin Coursey

Honors College Theses

Failure of physical devices can cause inconvenience, loss of money, and sometimes even deaths. To improve the reliability of these devices, we need to know the remaining useful life (RUL) of a device at a given point in time. Data-driven approaches use data from a physical device to build a model that can estimate the RUL. They have shown great performance and are often simpler than traditional model-based approaches. Typical statistical and machine learning approaches are often not suited for sequential data prediction. Recurrent Neural Networks are designed to work with sequential data but suffer from the vanishing gradient problem …