Computer Sciences Commons^™

Machine Learning And Network Embedding Methods For Gene Co-Expression Networks, Niloofar Aghaieabiane

Dissertations

High-throughput technologies such as DNA microarrays and RNA-seq are used to measure the expression levels of large numbers of genes simultaneously. To support the extraction of biological knowledge, individual gene expression levels are transformed into Gene Co-expression Networks (GCNs). GCNs are analyzed to discover gene modules. GCN construction and analysis is a well-studied topic, for nearly two decades. While new types of sequencing and the corresponding data are now available, the software package WGCNA and its most recent variants are still widely used, contributing to biological discovery.

The discovery of biologically significant modules of genes from raw expression data is …

Trustworthy Machine Learning Through The Lens Of Privacy And Security, Thi Kim Phung Lai

Dissertations

Nowadays, machine learning (ML) becomes ubiquitous and it is transforming society. However, there are still many incidents caused by ML-based systems when ML is deployed in real-world scenarios. Therefore, to allow wide adoption of ML in the real world, especially in critical applications such as healthcare, finance, etc., it is crucial to develop ML models that are not only accurate but also trustworthy (e.g., explainable, privacy-preserving, secure, and robust). Achieving trustworthy ML with different machine learning paradigms (e.g., deep learning, centralized learning, federated learning, etc.), and application domains (e.g., computer vision, natural language, human study, malware systems, etc.) is challenging, …

Mapping Programs To Equations, Hessamaldin Mohammadi

Dissertations

Extracting the function of a program from a static analysis of its source code is a valuable capability in software engineering; at a time when there is increasing talk of using AI (Artificial Intelligence) to generate software from natural language specifications, it becomes increasingly important to determine the exact function of software as written, to figure out what AI has understood the natural language specification to mean. For all its criticality, the ability to derive the domain-to-range function of a program has proved to be an elusive goal, due primarily to the difficulty of deriving the function of iterative statements. …

Deep Hybrid Modeling Of Neuronal Dynamics Using Generative Adversarial Networks, Soheil Saghafi

Dissertations

Mechanistic modeling and machine learning methods are powerful techniques for approximating biological systems and making accurate predictions from data. However, when used in isolation these approaches suffer from distinct shortcomings: model and parameter uncertainty limit mechanistic modeling, whereas machine learning methods disregard the underlying biophysical mechanisms. This dissertation constructs Deep Hybrid Models that address these shortcomings by combining deep learning with mechanistic modeling. In particular, this dissertation uses Generative Adversarial Networks (GANs) to provide an inverse mapping of data to mechanistic models and identifies the distributions of mechanistic model parameters coherent to the data.

Chapter 1 provides background information on …

Domain Decomposition Methods For Linear And Non-Linear Elliptic Problems, Tadanaga Takahashi

Dissertations

The primary purpose of this dissertation is to expand upon the circle of domain decomposition methods (DDM) which are algorithms that reformulate a boundary value problem in terms of multiple localized problems on subdomains. The first project involves expanding upon DDMs in a relatively mature field: the Helmholtz equation for wave scattering applications. The proposed method is an adaptation of a continuous cross-point Finite Element Non-overlapping DDM algorithm. The usual unbounded computational domain is truncated and then the near-field wave pattern is solved with a parallelized finite element method. Several improvements over the standard transmission operator are discussed in this …

A Survey On Online Matching And Ad Allocation, Ryan Lee

Theses

One of the classical problems in graph theory is matching. Given an undirected graph, find a matching which is a set of edges without common vertices. In 1990s, Richard Karp, Umesh Vazirani, and Vijay Vazirani would be the first computer scientists to use matchings for online algorithms [8]. In our domain, an online algorithm operates in the online setting where a bipartite graph is given. On one side of the graph there is a set of advertisers and on the other side we have a set of impressions. During the online phase, multiple impressions will arrive and the objective of …

Using Materialized Views For Answering Graph Pattern Queries, Michael Lan

Dissertations

Discovering patterns in graphs by evaluating graph pattern queries involving direct (edge-to-edge mapping) and reachability (edge-to-path mapping) relationships under homomorphisms on data graphs has been extensively studied. Previous studies have aimed to reduce the evaluation time of graph pattern queries due to the potentially numerous matches on large data graphs.

In this work, the concept of the summary graph is developed to improve the evaluation of tree pattern queries and graph pattern queries. The summary graph first filters out candidate matches which violate certain reachability constraints, and then finds local matches of query edges. This reduces redundancy in the representation …

Android Security: Analysis And Applications, Raina Samuel

Dissertations

The Android mobile system is home to millions of apps that offer a wide range of functionalities. Users rely on Android apps in various facets of daily life, including critical, e.g., medical, settings. Generally, users trust that apps perform their stated purpose safely and accurately. However, despite the platform’s efforts to maintain a safe environment, apps routinely manage to evade scrutiny. This dissertation analyzes Android app behavior and has revealed several weakness: lapses in device authentication schemes, deceptive practices such as apps covering their traces, as well as behavioral and descriptive inaccuracies in medical apps. Examining a large corpus of …

Machine Learning-Based Data Analytics For Understanding Space Weather And Climate, Yasser Abduallah

Dissertations

This dissertation addresses multiple crucial problems in space weather and climate, presenting new machine learning-based data analytics algorithms and models for tackling the problems.

First, the dissertation presents two new approaches to predicting solar flares. One approach, called DeepSun, predicts solar flares by utilizing a machine-learning-as-a-service (MLaaS) platform. The DeepSun system provides a friendly interface for Web users and an application programming interface (API) for remote programming users. It adopts an ensemble learning method that employs several machine learning algorithms to perform multiclass flare prediction. The other approach, named SolarFlareNet, forecasts the occurrence of solar flares within the next 24 …

Monitoring Climate Change With Machine Learning And Uncertainty Quantification, Firas Gerges

Dissertations

The impact of climate change on the environment has become increasingly visible today. One important topic in climate change monitoring is to forecast future climate events, which relies on long-term prediction of climate variables at a local regional scale. This topic is crucial for disaster risk management and for resolving climate change at the local level. General Circulation Models (GCMs) allow for the simulation of several climate variables, decades into the future. GCM simulations, however, are at a global large scale and are too coarse to monitor climate change at the local small scale. This dissertation explores new machine learning …

A Neural Analysis-Synthesis Approach To Learning Procedural Audio Models, Danzel Serrano

Theses

The effective sound design of environmental sounds is crucial to demonstrating an immersive experience. Classical Procedural Audio (PA) models have been developed to give the sound designer a fast way to synthesize a specific class of environmental sounds in a physically accurate and computationally efficient manner. These models are controllable due to the choice of parameters from analyzing a class of sound. However, the resulting synthesis lacks the fidelity for the preferred immersive experience; thus, the sound designer would rather search through an extensive database for real recordings of a target sound class. This thesis proposes the Procedural audio Variational …

Computation Of Risk Measures In Finance And Parallel Real-Time Scheduling, Yajuan Li

Dissertations

Many application areas employ various risk measures, such as a quantile, to assess risks. For example, in finance, risk managers employ a quantile to help determine appropriate levels of capital needed to be able to absorb (with high probability) large unexpected losses in credit portfolios comprising loans, bonds, and other financial instruments subject to default. This dissertation discusses the computation of risk measures in finance and parallel real-time scheduling.

Firstly, two estimation approaches are compared for one risk measure, a quantile, via randomized quasi-Monte Carlo (RQMC) in an asymptotic setting where the number of randomizations for RQMC grows large, but …

Low-Reynolds-Number Locomotion Via Reinforcement Learning, Yuexin Liu

Dissertations

This dissertation summarizes computational results from applying reinforcement learning and deep neural network to the designs of artificial microswimmers in the inertialess regime, where the viscous dissipation in the surrounding fluid environment dominates and the swimmer’s inertia is completely negligible. In particular, works in this dissertation consist of four interrelated studies of the design of microswimmers for different tasks: (1) a one-dimensional microswimmer in free-space that moves towards the target via translation, (2) a one-dimensional microswimmer in a periodic domain that rotates to reach the target, (3) a two-dimensional microswimmer that switches gaits to navigate to the designated targets in …

Towards Ensuring Integrity And Authenticity Of Software Repositories, Sangat Vaidya

Dissertations

The software development process comprises a series of steps known as a software supply chain. These steps include managing the source code, testing, building and packaging it into a final product, and distributing the product to end users. Along this chain, software repositories are used for different purposes such as source code management (Git, SVN, mercurial), software distribution (PyPI, RubyGems, NPM) or for deploying software based on container images (Harbor, DockerHub, Artifact Hub). In the recent past, different types of repositories have increasingly been the target of attacks. As such, there is a need for mechanisms to ensure integrity and …

Parameter Selection In Fully Homomorphic Encryption Schemes And Fhe Applications, Cavidan Yakupoglu

Dissertations

Quantum computing has been gaining momentum as a result of recent technological advances. Existing cryptographic systems rely on the difficult problems that can be solved by sufficiently powerful quantum computers. As the quantum age approaches, the desire to discover new difficult problems that cannot be solved by quantum systems has increased. Lattice-based cryptography is a prominent tool for the post-quantum era that facilitates the implementation of encryption systems for practical applications.

The Learning with Error (LWE) and Ring-LWE problems introduce new lattice hardness assumptions that have been incorporated into public-key cryptosystems to facilitate the implementation of numerous privacy-enhancing applications. Fully …

Efficient And Scalable Triangle Centrality Algorithms In The Arkouda Framework, Joseph Thomas Patchett

Theses

Graph data structures provide a unique challenge for both analysis and algorithm development. These data structures are irregular in that memory accesses are not known a priori and accesses to these structures tend to lack locality.

Despite these challenges, graph data structures are a natural way to represent relationships between entities and to exhibit unique features about these relationships. The network created from these relationships can create unique local structures that can describe the behavior between members of these structures. Graphs can be analyzed in a number of different ways including at a high level in community detection and at …

One-Stage Blind Source Separation Via A Sparse Autoencoder Framework, Jason Anthony Dabin

Dissertations

Blind source separation (BSS) is the process of recovering individual source transmissions from a received mixture of co-channel signals without a priori knowledge of the channel mixing matrix or transmitted source signals. The received co-channel composite signal is considered to be captured across an antenna array or sensor network and is assumed to contain sparse transmissions, as users are active and inactive aperiodically over time. An unsupervised machine learning approach using an artificial feedforward neural network sparse autoencoder with one hidden layer is formulated for blindly recovering the channel matrix and source activity of co-channel transmissions. The BSS sparse autoencoder …

Understanding The Voluntary Moderation Practices In Live Streaming Communities, Jie Cai

Dissertations

Harmful content, such as hate speech, online abuses, harassment, and cyberbullying, proliferates across various online communities. Live streaming as a novel online community provides ways for thousands of users (viewers) to entertain and engage with a broadcaster (streamer) in real-time in the chatroom. While the streamer has the camera on and the screen shared, tens of thousands of viewers are watching and messaging in real-time, resulting in concerns about harassment and cyberbullying. To regulate harmful content—toxic messages in the chatroom, streamers rely on a combination of automated tools and volunteer human moderators (mods) to block users or remove content, which …

A Self-Learning Intersection Control System For Connected And Automated Vehicles, Ardeshir Mirbakhsh

Dissertations

This study proposes a Decentralized Sparse Coordination Learning System (DSCLS) based on Deep Reinforcement Learning (DRL) to control intersections under the Connected and Automated Vehicles (CAVs) environment. In this approach, roadway sections are divided into small areas; vehicles try to reserve their desired area ahead of time, based on having a common desired area with other CAVs; the vehicles would be in an independent or coordinated state. Individual CAVs are set accountable for decision-making at each step in both coordinated and independent states. In the training process, CAVs learn to minimize the overall delay at the intersection. Due to the …

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 …

Optimization Opportunities In Human In The Loop Computational Paradigm, Dong Wei

Dissertations

An emerging trend is to leverage human capabilities in the computational loop at different capacities, ranging from tapping knowledge from a richly heterogeneous pool of knowledge resident in the general population to soliciting expert opinions. These practices are, in general, termed human-in-the-loop (HITL) computations.

A HITL process requires holistic treatment and optimization from multiple standpoints considering all stakeholders: a. applications, b. platforms, c. humans. In application-centric optimization, the factors of interest usually are latency (how long it takes for a set of tasks to finish), cost (the monetary or computational expenses incurred in the process), and quality of the completed …

Towards Practicalization Of Blockchain-Based Decentralized Applications, Songlin He

Dissertations

Blockchain can be defined as an immutable ledger for recording transactions, maintained in a distributed network of mutually untrusting peers. Blockchain technology has been widely applied to various fields beyond its initial usage of cryptocurrency. However, blockchain itself is insufficient to meet all the desired security or efficiency requirements for diversified application scenarios. This dissertation focuses on two core functionalities that blockchain provides, i.e., robust storage and reliable computation. Three concrete application scenarios including Internet of Things (IoT), cybersecurity management (CSM), and peer-to-peer (P2P) content delivery network (CDN) are utilized to elaborate the general design principles for these two main …

Representation Learning In Finance, Ajim Uddin

Dissertations

Finance studies often employ heterogeneous datasets from different sources with different structures and frequencies. Some data are noisy, sparse, and unbalanced with missing values; some are unstructured, containing text or networks. Traditional techniques often struggle to combine and effectively extract information from these datasets. This work explores representation learning as a proven machine learning technique in learning informative embedding from complex, noisy, and dynamic financial data. This dissertation proposes novel factorization algorithms and network modeling techniques to learn the local and global representation of data in two specific financial applications: analysts’ earnings forecasts and asset pricing.

Financial analysts’ earnings forecast …

Towards A Cross-Layer Coupled Design Framework For Big Data Workflows, Qianwen Ye

Dissertations

The processing and analysis of big data increasingly rely on workflow technologies for knowledge discovery and scientific innovation. The execution of such workflows goes far beyond the capability and capacity of single computers and is now commonly supported on reliable and scalable data storage and analysis platforms in distributed environments, such as the Hadoop ecosystem. Workflow performance largely depends on how big data systems are configured and used. For example, the makespan of a big data workflow is affected by multiple layers of big data systems, including the parallel computing engine it runs on, the resource manager that orchestrates various …

Graph Enabled Cross-Domain Knowledge Transfer, Shibo Yao

Dissertations

The world has never been more connected, led by the information technology revolution in the past decades that has fundamentally changed the way people interact with each other using social networks. Consequently, enormous human activity data are collected from the business world and machine learning techniques are widely adopted to aid our decision processes. Despite of the success of machine learning in various application scenarios, there are still many questions that need to be well answered, such as optimizing machine learning outcomes when desired knowledge cannot be extracted from the available data. This naturally drives us to ponder if one …

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 …

Private Information Retrieval And Function Computation For Noncolluding Coded Databases, Sarah A. Obead

Dissertations

The rapid development of information and communication technologies has motivated many data-centric paradigms such as big data and cloud computing. The resulting paradigmatic shift to cloud/network-centric applications and the accessibility of information over public networking platforms has brought information privacy to the focal point of current research challenges. Motivated by the emerging privacy concerns, the problem of private information retrieval (PIR), a standard problem of information privacy that originated in theoretical computer science, has recently attracted much attention in the information theory and coding communities. The goal of PIR is to allow a user to download a message from a …

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 …

Design And Implementation Of Photovoltaic Energy Harvesting Automaton, Iskandar Askarov

Theses

Global domestic electricity consumption has been rapidly increasing in the past three decades. In fact, from 1990 to 2020, consumption has more than doubled from 10,120 TWh to 23,177 TWh [1]. Moreover, consumers have been turning more towards clean, renewable energy sources such as Photovoltaic. According to International Energy Agency, global Solar power generation alone in 2019 has reached almost 3% [4] of the electricity supply. Even though the efficiency of photovoltaic panels has been growing, presently, the highest efficiency solar panels available to an average consumer range only from 20%-22% [14]. Many research papers have been published to increase …

Exposing And Fixing Causes Of Inconsistency And Nondeterminism In Clustering Implementations, Xin Yin

Dissertations

Cluster analysis aka Clustering is used in myriad applications, including high-stakes domains, by millions of users. Clustering users should be able to assume that clustering implementations are correct, reliable, and for a given algorithm, interchangeable. Based on observations in a wide-range of real-world clustering implementations, this dissertation challenges the aforementioned assumptions.

This dissertation introduces an approach named SmokeOut that uses differential clustering to show that clustering implementations suffer from nondeterminism and inconsistency: on a given input dataset and using a given clustering algorithm, clustering outcomes and accuracy vary widely between (1) successive runs of the same toolkit, i.e., nondeterminism, and …