Engineering Commons^™

Sc-Fuse: A Feature Fusion Approach For Unpaved Road Detection From Remotely Sensed Images, Aniruddh Saxena

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Road network extraction from remote sensing imagery is crucial for numerous applications, ranging from autonomous navigation to urban and rural planning. A particularly challenging aspect is the detection of unpaved roads, often underrepresented in research and data. These roads display variability in texture, width, shape, and surroundings, making their detection quite complex. This thesis addresses these challenges by creating a specialized dataset and introducing the SC-Fuse model.

Our custom dataset comprises high resolution remote sensing imagery which primarily targets unpaved roads of the American Midwest. To capture the diverse seasonal variation and their impact, the dataset includes images from different …

Experimental Study Of Linux Flightsize Estimation, Mingrui Zhang

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Transmission Control Protocol (TCP) is a fundamental Internet protocol responsible for controlling and coordinating the Internet traffic. As a result, TCP significantly influences the overall performance and stability of the Internet. One critical information required by a TCP connection to make decisions is FlightSize, which is the total amount of outstanding data contributed by the connection to the Internet. The FlightSize information is used by a TCP connection to determine its future sending rate and also avoid traffic congestion and collapse in the Internet. Consequently, an inaccurate estimation of FlightSize can result in degraded performance and instability of the Internet. …

Sim-To-Real Reinforcement Learning Framework For Autonomous Aerial Leaf Sampling, Ashraful Islam

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Using unmanned aerial systems (UAS) for leaf sampling is contributing to a better understanding of the influence of climate change on plant species, and the dynamics of forest ecology by studying hard-to-reach tree canopies. Currently, multiple skilled operators are required for UAS maneuvering and using the leaf sampling tool. This often limits sampling to only the canopy top or periphery. Sim-to-real reinforcement learning (RL) can be leveraged to tackle challenges in the autonomous operation of aerial leaf sampling in the changing environment of a tree canopy. However, trans- ferring an RL controller that is learned in simulation to real UAS …

Leveraging Aruco Fiducial Marker System For Bridge Displacement Estimation Using Unmanned Aerial Vehicles, Mohamed Aly

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The use of unmanned aerial vehicles (UAVs) in construction sites has been widely growing for surveying and inspection purposes. Their mobility and agility have enabled engineers to use UAVs in Structural Health Monitoring (SHM) applications to overcome the limitations of traditional approaches that require labor-intensive installation, extended time, and long-term maintenance. One of the critical applications of SHM is measuring bridge deflections during the bridge operation period. Due to the complex remote sites of bridges, remote sensing techniques, such as camera-equipped drones, can facilitate measuring bridge deflections. This work takes a step to build a pipeline using the state-of-the-art computer …

An Empirical Study On The Classification Of Python Language Features Using Eye-Tracking, Jigyasa Chauhan

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Python, currently one of the most popular programming languages, is an object-
oriented language that also provides language feature support for other programming
paradigms, such as functional and procedural. It is not currently understood how
support for multiple paradigms affects the ability of developers to comprehend that
code. Understanding the predominant paradigm in code, and how developers classify
the predominant paradigm, can benefit future research in program comprehension as
the paradigm may factor into how people comprehend that code. Other researchers
may want to look at how the paradigms in the code interact with various code smells.
To investigate how …

Attention In The Faithful Self-Explanatory Nlp Models, Mostafa Rafaiejokandan

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Deep neural networks (DNNs) can perform impressively in many natural language processing (NLP) tasks, but their black-box nature makes them inherently challenging to explain or interpret. Self-Explanatory models are a new approach to overcoming this challenge, generating explanations in human-readable languages besides task objectives like answering questions. The main focus of this thesis is the explainability of NLP tasks, as well as how attention methods can help enhance performance. Three different attention modules are proposed, SimpleAttention, CrossSelfAttention, and CrossModality. It also includes a new dataset transformation method called Two-Documents that converts every dataset into two separate documents required by the …

Learnfca: A Fuzzy Fca And Probability Based Approach For Learning And Classification, Suraj Ketan Samal

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Formal concept analysis(FCA) is a mathematical theory based on lattice and order theory used for data analysis and knowledge representation. Over the past several years, many of its extensions have been proposed and applied in several domains including data mining, machine learning, knowledge management, semantic web, software development, chemistry ,biology, medicine, data analytics, biology and ontology engineering.

This thesis reviews the state-of-the-art of theory of Formal Concept Analysis(FCA) and its various extensions that have been developed and well-studied in the past several years. We discuss their historical roots, reproduce the original definitions and derivations with illustrative examples. Further, we provide …

Bevers: A General, Simple, And Performant Framework For Automatic Fact Verification, Mitchell Dehaven

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Fact verification has become an important process, primarily done manually by humans, to verify the authenticity of claims and statements made online. Increasingly, social media companies have utilized human effort to debunk false claims on their platforms, opting to either tag the content as misleading or false, or removing it entirely to combat misinformation on their sites. In tandem, the field of automatic fact verification has become a subject of focus among the natural language processing (NLP) community, spawning new datasets and research. The most popular dataset is the Fact Extraction and VERification (FEVER) dataset. In this thesis an end-to-end …

Sequence-Based Bioinformatics Approaches To Predict Virus-Host Relationships In Archaea And Eukaryotes, Yingshan Li

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Viral metagenomics is independent of lab culturing and capable of investigating viromes of virtually any given environmental niches. While numerous sequences of viral genomes have been assembled from metagenomic studies over the past years, the natural hosts for the majority of these viral contigs have not been determined. Different computational approaches have been developed to predict hosts of bacteria phages. Nevertheless, little progress has been made in the virus-host prediction, especially for viruses that infect eukaryotes and archaea. In this study, by analyzing all documented viruses with known eukaryotic and archaeal hosts, we assessed the predictive power of four computational …

A Pipeline To Generate Deep Learning Surrogates Of Genome-Scale Metabolic Models, Achilles Rasquinha

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Genome-Scale Metabolic Models (GEMMs) are powerful reconstructions of biological systems that help metabolic engineers understand and predict growth conditions subjected to various environmental factors around the cellular metabolism of an organism in observation, purely in silico. Applications of metabolic engineering range from perturbation analysis and drug-target discovery to predicting growth rates of biotechnologically important metabolites and reaction objectives within dierent single-cell and multi-cellular organism types. GEMMs use mathematical frameworks for quantitative estimations of flux distributions within metabolic networks. The reasons behind why an organism activates, stuns, or fluctuates between alternative pathways for growth and survival, however, remain relatively unknown. GEMMs …

Feed Forward Neural Networks With Asymmetric Training, Archit Srivastava

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Our work presents a new perspective on training feed-forward neural networks(FFNN). We introduce and formally define the notion of symmetry and asymmetry in the context of training of FFNN. We provide a mathematical definition to generalize the idea of sparsification and demonstrate how sparsification can induce asymmetric training in FFNN.

In FFNN, training consists of two phases, forward pass and backward pass. We define symmetric training in FFNN as follows-- If a neural network uses the same parameters for both forward pass and backward pass, then the training is said to be symmetric.

The definition of asymmetric training in artificial …

Simulating Sub-Threshold Communication Channels Through Neurons, Richard Maina

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Molecular Communication is an emerging paradigm with the potential to revolutionize the technology behind wearable and implantable devices and the broad range of functions they support, from tracking physical activity to medical diagnostics. This can be achieved through intra-body communication networks that take advantage of natural biological processes as a means of transmitting, propagating and receiving information. In this thesis we focus particularly on using the neuron as a means to facilitate information transfer for interconnected wearable or implantable devices through a technique known as sub-threshold electrical stimulation. We develop upon a prior work by introducing a linear model of …

Consemblex: A Consensus-Based Transcriptome Assembly Approach That Extends Consemble And Improves Transcriptome Assembly, Richard Mwaba

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

An accurate transcriptome is essential to understanding biological systems enabling omics analyses such as gene expression, gene discovery, and gene-regulatory network construction. However, assembling an accurate transcriptome is challenging, especially for organisms without adequate reference genomes or transcriptomes. While several methods for transcriptome assembly with different approaches exist, it is still difficult to establish the most accurate methods. This thesis explores the different transcriptome assembly methods and compares their performances using simulated benchmark transcriptomes with varying complexity. We also introduce ConSemblEX to improve a consensus-based ensemble transcriptome assembler, ConSemble, in three main areas: we provide the ability to use any …

Symbolic Ns-3 For Efficient Exhaustive Testing, Jianfei Shao

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Exhaustive testing is an important type of simulation, where a user exhaustively simulates a protocol for all possible cases with respect to some uncertain factors, such as all possible packet delays or packet headers. It is useful for completely evaluating the protocol performance, finding the worst-case performance, and detecting possible design or implementation bugs of a protocol. It is, however, time consuming to use the brute force method with current NS-3, a widely used network simulator, for exhaustive testing. In this paper, we present our work on Sym-NS-3 for more efficient exhaustive testing, which leverages a powerful program analysis technique …

Machine Learning-Based Device Type Classification For Iot Device Re- And Continuous Authentication, Kaustubh Gupta

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Today, the use of Internet of Things (IoT) devices is higher than ever and it is growing rapidly. Many IoT devices are usually manufactured by home appliance manufacturers where security and privacy are not the foremost concern. When an IoT device is connected to a network, currently there does not exist a strict authentication method that verifies the identity of the device, allowing any rogue IoT device to authenticate to an access point. This thesis addresses the issue by introducing methods for continuous and re-authentication of static and dynamic IoT devices, respectively. We introduce mechanisms and protocols for authenticating a …

Characterizing And Predicting Human Visual Perception Of Unmanned Aerial Vehicle Gestures, Paul Fletcher

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Unmanned Aerial Vehicles (UAVs) are being used in public domains and hazardous environments where effective communication strategies are critical. UAV gesture techniques have been shown to communicate meaning to human observers and may be ideal in contexts that require lightweight systems such as unmanned aerial flight, however, this work may be limited to an idealized range of viewer perspectives. As gesture is a visual communication technique it is necessary to consider how the perception of a robot gesture may suffer from obfuscation or self-occlusion from some viewpoints. This thesis presents the results of three online user-studies that examine participants’ ability …

Computational Solutions To Exosomal Microrna Biomarker Detection In Pancreatic Cancer, Thuy T. An

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Pancreatic cancer is the fourth leading cause of cancer death in the United States and the 5-year survival rate is only 5% to 10%. There are only a few non-specific symptoms associated with the early-stage cancer, therefore most patients are diagnosed in a late stage. Due to the lack of effective treatments and the fact that the early stage has a 39% 5-year survival rate, the biggest hope to control this disease is early detection. Therefore, discovery of effective and reliable non-invasive biomarkers for early detection of pancreatic cancer has been a major topic. Very recently, exosomal microRNAs have become …

Semantically Meaningful Sentence Embeddings, Rojina Deuja

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Text embedding is an approach used in Natural Language Processing (NLP) to represent words, phrases, sentences, and documents. It is the process of obtaining numeric representations of text to feed into machine learning models as vectors (arrays of numbers). One of the biggest challenges in text embedding is representing longer text segments like sentences. These representations should capture the meaning of the segment and the semantic relationship between its constituents. Such representations are known as semantically meaningful embeddings. In this thesis, we seek to improve upon the quality of sentence embeddings that capture semantic information.

The current state-of-the-art models are …

Comparative Analysis Of Kmer Counting And Estimation Tools, Ankitha Vejandla

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The rapid development of next-generation sequencing (NGS) technologies for determining the sequence of DNA has revolutionized genome research in recent years. De novo assemblers are the most commonly used tools to perform genome assembly. Most of the assemblers use de Bruijn graphs that break the sequenced reads into smaller sequences (sub-strings), called kmers, where k denotes the length of the sub-strings. The kmer counting and analysis of kmer frequency distribution are important in genome assembly. The main goal of this research is to provide a detailed analysis of the performance of different kmer counting and estimation tools that are currently …

Power-Over-Tether Unmanned Aerial System Leveraged For Trajectory Influenced Atmospheric Sensing, Daniel Rico

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The use of unmanned aerial systems (UASs) in agriculture has risen in the past decade and is helping to modernize agriculture. UASs collect and elucidate data previously difficult to obtain and are used to help increase agricultural efficiency and production. Typical commercial off-the-shelf (COTS) UASs are limited by small payloads and short flight times. Such limits inhibit their ability to provide abundant data at multiple spatiotemporal scales. In this thesis, we describe the design and construction of the tethered aircraft unmanned system (TAUS), which is a novel power-over-tether UAS configured for long-term, high throughput atmospheric monitoring with an array of …

A Real-World, Hybrid Event Sequence Generation Framework For Android Apps, Jun Sun

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Generating meaningful inputs for Android apps is still a challenging issue that needs more research. Past research efforts have shown that random test generation is still an effective means to exercise User-Interface (UI) events to achieve high code coverage. At the same time, heuristic search approaches can effectively reach specified code targets. Our investigation shows that these approaches alone are insufficient to generate inputs that can exercise specific code locations in complex Android applications.

This thesis introduces a hybrid approach that combines two different input generation techniques--heuristic search based on genetic algorithm and random instigation of UI events, to reach …

Using Contextual Bandits To Improve Traffic Performance In Edge Network, Aziza Al Zadjali

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Edge computing network is a great candidate to reduce latency and enhance performance of the Internet. The flexibility afforded by Edge computing to handle data creates exciting range of possibilities. However, Edge servers have some limitations since Edge computing process and analyze partial sets of information. It is challenging to allocate computing and network resources rationally to satisfy the requirement of mobile devices under uncertain wireless network, and meet the constraints of datacenter servers too. To combat these issues, this dissertation proposes smart multi armed bandit algorithms that decide the appropriate connection setup for multiple network access technologies on the …

Aerial Flight Paths For Communication, Alisha Bevins

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

This body of work presents an iterative process of refinement to understand naive perception of communication using the motion of an unmanned aerial vehicle (UAV). This includes what people believe the UAV is trying to communicate, and how they expect to respond through physical action or emotional response. Previous work in this area sought to communicate without clear definitions of the states attempting to be conveyed. In an attempt to present more concrete states and better understand specific motion perception, this work goes through multiple iterations of state elicitation and label assignment. The lessons learned in this work will be …

Using An Integrative Machine Learning Approach To Study Microrna Regulation Networks In Pancreatic Cancer Progression, Roland Madadjim

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

With advances in genomic discovery tools, recent biomedical research has produced a massive amount of genomic data on post-transcriptional regulations related to various transcript factors, microRNAs, lncRNAs, epigenetic modifications, and genetic variations. In this direction, the field of gene regulation network inference is created and aims to understand the interactome regulations between these molecules (e.g., gene-gene, miRNA-gene) that take place to build models able to capture behavioral changes in biological systems. A question of interest arises in integrating such molecules to build a network while treating each specie in its uniqueness. Given the dynamic changes of interactome in chaotic systems …

Teachability And Interpretability In Reinforcement Learning, Jeevan Rajagopal

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

There have been many recent advancements in the field of reinforcement learning, starting from the Deep Q Network playing various Atari 2600 games all the way to Google Deempind's Alphastar playing competitively in the game StarCraft. However, as the field challenges more complex environments, the current methods of training models and understanding their decision making become less effective. Currently, the problem is partially dealt with by simply adding more resources, but the need for a better solution remains.

This thesis proposes a reinforcement learning framework where a teacher or entity with domain knowledge of the task to complete can assist …

“The Revolution Will Not Be Supervised": An Investigation Of The Efficacy And Reasoning Process Of Self-Supervised Representations, Atharva Tendle

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Transfer learning technique enables training Deep Learning (DL) models in a data-efficient way for solving computer vision tasks. It involves pretraining a DL model to learn representations from a large and general-purpose source dataset, then fine-tuning the model using the task-specific target dataset. The dominant supervised learning (SL) approach for pretraining representations suffers from some limitations that include expensive labeling and poor generalizability. Recent advancements in the self-supervised learning (SSL) approach made it possible to learn effective representations from unlabeled data. The performance of the fine-tuned DL models based on pretrained SSL representations is on par with the state-of-the-art pretrained …

A Novel Spatiotemporal Prediction Method Of Cumulative Covid-19 Cases, Junzhe Cai

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Prediction methods are important for many applications. In particular, an accurate prediction for the total number of cases for pandemics such as the Covid-19 pandemic could help medical preparedness by providing in time a sufficient supply of testing kits, hospital beds and medical personnel. This thesis experimentally compares the accuracy of ten prediction methods for the cumulative number of Covid-19 pandemic cases. These ten methods include two types of neural networks and extrapolation methods based on best fit linear, best fit quadratic, best fit cubic and Lagrange interpolation, as well as an extrapolation method from Revesz. We also consider the …

Suffix Tree, Minwise Hashing And Streaming Algorithms For Big Data Analysis In Bioinformatics, Sairam Behera

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

In this dissertation, we worked on several algorithmic problems in bioinformatics using mainly three approaches: (a) a streaming model, (b) sux-tree based indexing, and (c) minwise-hashing (minhash) and locality-sensitive hashing (LSH). The streaming models are useful for large data problems where a good approximation needs to be achieved with limited space usage. We developed an approximation algorithm (Kmer-Estimate) using the streaming approach to obtain a better estimation of the frequency of k-mer counts. A k-mer, a subsequence of length k, plays an important role in many bioinformatics analyses such as genome distance estimation. We also developed new methods that use …

Representational Learning Approach For Predicting Developer Expertise Using Eye Movements, Sumeet Maan

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

The thesis analyzes an existing eye-tracking dataset collected while software developers were solving bug fixing tasks in an open-source system. The analysis is performed using a representational learning approach namely, Multi-layer Perceptron (MLP). The novel aspect of the analysis is the introduction of a new feature engineering method based on the eye-tracking data. This is then used to predict developer expertise on the data. The dataset used in this thesis is inherently more complex because it is collected in a very dynamic environment i.e., the Eclipse IDE using an eye-tracking plugin, iTrace. Previous work in this area only worked on …

Formal Concept Analysis Applications In Bioinformatics, Sarah Roscoe

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Bioinformatics is an important field that seeks to solve biological problems with the help of computation. One specific field in bioinformatics is that of genomics, the study of genes and their functions. Genomics can provide valuable analysis as to the interaction between how genes interact with their environment. One such way to measure the interaction is through gene expression data, which determines whether (and how much) a certain gene activates in a situation. Analyzing this data can be critical for predicting diseases or other biological reactions. One method used for analysis is Formal Concept Analysis (FCA), a computing technique based …