Physical Sciences and Mathematics Commons^™

Interactive Data Analysis Of Multi-Run Performance Data, Vanessa Lama

Masters Theses

Multi-dimensional performance data analysis presents challenges for programmers, and users. Developers have to choose library and compiler options for each platform, analyze raw performance data, and keep up with new technologies. Users run codes on different platforms, validate results with collaborators, and analyze performance data as applications scale up. Site operators use multiple profiling tools to optimize performance, requiring the analysis of multiple sources and data types. There is currently no comprehensive tool to support the structured analysis of unstructured data, when holistic performance data analysis can offer actionable insights and improve performance. In this work, we present thicket, a …

A Machine Learning Approach For Predicting Clinical Trial Patient Enrollment In Drug Development Portfolio Demand Planning, Ahmed Shoieb

Masters Theses

One of the biggest challenges the clinical research industry currently faces is the accurate forecasting of patient enrollment (namely if and when a clinical trial will achieve full enrollment), as the stochastic behavior of enrollment can significantly contribute to delays in the development of new drugs, increases in duration and costs of clinical trials, and the over- or under- estimation of clinical supply. This study proposes a Machine Learning model using a Fully Convolutional Network (FCN) that is trained on a dataset of 100,000 patient enrollment data points including patient age, patient gender, patient disease, investigational product, study phase, blinded …

Survey Of Input Modalities In The Western World, John Ezat Sadik

Masters Theses

Having your account compromised can lead to serious complications in your life. One
way accounts become compromised is through the security risks associated with weak
passwords and reused passwords [22,23]. In this thesis, we seek to understand how
entering passwords on non-PC devices contributes to the problems of weak and reused
passwords. To do so, we conducted a survey that was distributed to people in the
the Western World. In our survey results, we found that users commented about
how the current password model was not created with a variety of device types in
mind, which created frustrations and complexity …

Meta-Heuristic Optimization Techniques For The Production Of Medical Isotopes Through Special Target Design, Cameron Ian Salyer

Masters Theses

Medical isotopes are used for a variety of different diagnostic and therapeutic purposes Ruth (2008). Due to recent newly discovered applications, their production has become rapidly more scarce than ever before Charlton (2019). Therefore, more efficient and less time consuming methods are of interest for not only the industry’s demand, but for the individuals who require radio-isotope procedures. Currently, the primary source of most medical isotopes used today are provided by reactor and cyclotron irradiation techniques, followed by supplemental radio-chemical separations Ruth (2008). Up until this point, target designs have been optimized by experience, back of the envelope calculations, and …

Accelerating Dynamical Density Response Code On Summit And Its Application For Computing The Density Response Function Of Vanadium Sesquioxide, Wileam Y. Phan

Masters Theses

This thesis details the process of porting the Eguiluz group dynamical density response computational platform to the hybrid CPU+GPU environment at the Summit supercomputer at Oak Ridge National Laboratory (ORNL) Leadership Computing Center. The baseline CPU-only version is a Gordon Bell-winning platform within the formally-exact time-dependent density functional theory (TD-DFT) framework using the linearly augmented plane wave (LAPW) basis set. The code is accelerated using a combination of the OpenACC programming model and GPU libraries -- namely, the Matrix Algebra for GPU and Multicore Architectures (MAGMA) library -- as well as exploiting the sparsity pattern of the matrices involved in …

Evaluation Of Robust Deep Learning Pipelines Targeting Low Swap Edge Deployment, David Carter Cornett

Masters Theses

The deep learning technique of convolutional neural networks (CNNs) has greatly advanced the state-of-the-art for computer vision tasks such as image classification and object detection. These solutions rely on large systems leveraging wattage-hungry GPUs to provide the computational power to achieve such performance. However, the size, weight and power (SWaP) requirements of these conventional GPU-based deep learning systems are not suitable when a solution requires deployment to so called "Edge" environments such as autonomous vehicles, unmanned aerial vehicles (UAVs) and smart security cameras.

The objective of this work is to benchmark FPGA-based alternatives to conventional GPU systems that have the …

Power System Stability Assessment With Supervised Machine Learning, Mirka Mandich

Masters Theses

Power system stability assessment has become an important area of research due to the increased penetration of photovoltaics (PV) in modern power systems. This work explores how supervised machine learning can be used to assess power system stability for the Western Electricity Coordinating Council (WECC) service region as part of the Data-driven Security Assessment for the Multi-Timescale Integrated Dynamics and Scheduling for Solar (MIDAS) project. Data-driven methods offer to improve power flow scheduling through machine learning prediction, enabling better energy resource management and reducing demand on real-time time-domain simulations. Frequency, transient, and small signal stability datasets were created using the …

Random Search Plus: A More Effective Random Search For Machine Learning Hyperparameters Optimization, Bohan Li

Masters Theses

Machine learning hyperparameter optimization has always been the key to improve model performance. There are many methods of hyperparameter optimization. The popular methods include grid search, random search, manual search, Bayesian optimization, population-based optimization, etc. Random search occupies less computations than the grid search, but at the same time there is a penalty for accuracy. However, this paper proposes a more effective random search method based on the traditional random search and hyperparameter space separation. This method is named random search plus. This thesis empirically proves that random search plus is more effective than random search. There are some case …

Information Theory Problem Description Parser, Gary Brent Hurst

Masters Theses

Data corruption and data loss create huge problems when they occur, so naturally safeguards are usually in place to recover lost data. This often involves allowing less space for data in order to allow space for an encoding that can be used to recover any data that might be lost. The question arises, then, about how to most efficiently implement these safeguards with respect to storage, network bandwidth, or some linear combination of those two things. This work has two main goals for the information theory community: to produce an intuitive-to-use problem description parser that facilitates research in the area, …

Improving Convolutional Neural Network Robustness To Adversarial Images Through Image Filtering, Natalie E. Bogda

Masters Theses

The field of computer vision and deep learning is known for its ability to recognize images with extremely high accuracy. Convolutional neural networks exist that can correctly classify 96\% of 1.2 million images of complex scenes. However, with just a few carefully positioned imperceptible changes to the pixels of an input image, an otherwise accurate network will misclassify this almost identical image with high confidence. These perturbed images are known as \textit{adversarial examples} and expose that convolutional neural networks do not necessarily "see" the world in the way that humans do. This work focuses on increasing the robustness of classifiers …

A Privacy Evaluation Of Nyx, Savannah A. Norem

Masters Theses

For this project, I will be analyzing the privacy leakage in a certain DDoS mitigation system. Nyx has been shown both in simulation and over live internet traffic to mitigate the effects of DDoS without any cooperation from downstream ASes and without any modifications to current routing protocols. However it does this through BPG-poisoning, which can unintentionally advertise information. This project explores what the traffic from Nyx looks like and what information can be gathered from it. Specifically, Nyx works by defining a deployer/critical relationship whose traffic is moved to maintain even under DDoS circumstances, and I will be evaluating …

Peer Attention Modeling With Head Pose Trajectory Tracking Using Temporal Thermal Maps, Corey Michael Johnson

Masters Theses

Human head pose trajectories can represent a wealth of implicit information such as areas of attention, body language, potential future actions, and more. This signal is of high value for use in Human-Robot teams due to the implicit information encoded within it. Although team-based tasks require both explicit and implicit communication among peers, large team sizes, noisy environments, distance, and mission urgency can inhibit the frequency and quality of explicit communication. The goal for this thesis is to improve the capabilities of Human-Robot teams by making use of implicit communication. In support of this goal, the following hypotheses are investigated: …

Developing Leading And Lagging Indicators To Enhance Equipment Reliability In A Lean System, Dhanush Agara Mallesh

Masters Theses

With increasing complexity in equipment, the failure rates are becoming a critical metric due to the unplanned maintenance in a production environment. Unplanned maintenance in manufacturing process is created issues with downtimes and decreasing the reliability of equipment. Failures in equipment have resulted in the loss of revenue to organizations encouraging maintenance practitioners to analyze ways to change unplanned to planned maintenance. Efficient failure prediction models are being developed to learn about the failures in advance. With this information, failures predicted can reduce the downtimes in the system and improve the throughput.

The goal of this thesis is to predict …

Automated Program Profiling And Analysis For Managing Heterogeneous Memory Systems, Adam Palmer Howard

Masters Theses

Many promising memory technologies, such as non-volatile, storage-class memories and high-bandwidth, on-chip RAMs, are beginning to emerge. Since each of these new technologies present tradeoffs distinct from conventional DRAMs, next-generation systems are likely to include multiple tiers of memory storage, each with their own type of devices. To efficiently utilize the available hardware, such systems will need to alter their data management strategies to consider the performance and capabilities provided by each tier.

This work explores a variety of cross-layer strategies for managing application data in heterogeneous memory systems. We propose different program profiling-based techniques to automatically partition program allocation …

The Synthesis Of Memristive Neuromorphic Circuits, Austin Richard Wyer

Masters Theses

As Moores Law has come to a halt, it has become necessary to explore alternative forms of computation that are not limited in the same ways as traditional CMOS technologies and the Von Neumann architecture. Neuromorphic computing, computing inspired by the human brain with neurons and synapses, has been proposed as one of these alternatives. Memristors, non-volatile devices with adjustable resistances, have emerged as a candidate for implementing neuromorphic computing systems because of their low power and low area overhead. This work presents a C++ simulator for an implementation of a memristive neuromorphic circuit. The simulator is used within a …

A Gpu Implementation Of Distance-Driven Computed Tomography, Ryan D. Wagner

Masters Theses

Computed tomography (CT) is used to produce cross-sectional images of an object via noninvasive X-ray scanning of the object. These images have a wide range of uses including threat detection in checked baggage at airports. The projection data collected by the CT scanner must be reconstructed before the image may be viewed. In comparison to filtered backprojection methods of reconstruction, iterative reconstruction algorithms have been shown to increase overall image quality by incorporating a more complete model of the underlying physics. Unfortunately, iterative algorithms are generally too slow to meet the high throughput demands of this application. It is therefore …

On The Role Of Genetic Algorithms In The Pattern Recognition Task Of Classification, Isaac Ben Sherman

Masters Theses

In this dissertation we ask, formulate an apparatus for answering, and answer the following three questions: Where do Genetic Algorithms fit in the greater scheme of pattern recognition? Given primitive mechanics, can Genetic Algorithms match or exceed the performance of theoretically-based methods? Can we build a generic universal Genetic Algorithm for classification? To answer these questions, we develop a genetic algorithm which optimizes MATLAB classifiers and a variable length genetic algorithm which does classification based entirely on boolean logic. We test these algorithms on disparate datasets rooted in cellular biology, music theory, and medicine. We then get results from these …

Efficient Simulation Of A Simple Evolutionary System, Mahendra Duwal Shrestha

Masters Theses

An infinite population model is considered for diploid evolution under the influence of crossing over and mutation. The evolution equations show how Vose’s haploid model for Genetic Algorithms extends to the diploid case, thereby making feasible simulations which otherwise would require excessive resources. This is illustrated through computations confirming the convergence of finite diploid population short-term behaviour to the behaviour predicted by the infinite diploid model. The results show the distance between finite and infinite population evolutionary trajectories can decrease in practice like the reciprocal of the square root of population size.

Under necessary and sufficient conditions (NS) concerning mutation …

Tagamajig: Image Recognition Via Crowdsourcing, Gregory Martin Simpson

Masters Theses

The University of Tennessee, Knoxville (UTK) Library possesses thousands of unlabeled gray-scale photographs from the Smoky Mountains circa the 1920s - 1940s. Their current method of identifying and labeling attributes of the photographs is to do so manually. This is problematic both because of the scale of the collection as well as the reliance on an individual's limited knowledge of the area's numerous landmarks.

In the past few years, similar dilemmas have been tackled via an approach known as crowd computing. Some examples include Floating Forests, in which users are asked to identify and mark kelp forests in satellite images, …

Three Body Interactions Of Rare Gas Solids Calculated Within The Einstein Model, Dan D'Andrea

Masters Theses

Three body interactions can become important in solids at higher pressures and densities as the molecules can come into close contact. At low temperatures, accurate studies of three body interactions in solids require averaging the three-body terms over the molecules' zero point motions. An efficient, but approximate, averaging approach is based on a polynomial approximation of the three-body term. The polynomial approximation can be developed as a function of the symmetry coordinates of a triangle displaced from its average geometry and also as a function of the Cartesian zero point displacements from each atom’s average position. The polynomial approximation approach …

Pdroid, Joe Larry Allen

Masters Theses

When an end user attempts to download an app on the Google Play Store they receive two related items that can be used to assess the potential threats of an application, the list of permissions used by the application and the textual description of the application. However, this raises several concerns. First, applications tend to use more permissions than they need and end users are not tech-savvy enough to fully understand the security risks. Therefore, it is challenging to assess the threats of an application fully by only seeing the permissions. On the other hand, most textual descriptions do not …

Hsp-Wrap: The Design And Evaluation Of Reusable Parallelism For A Subclass Of Data-Intensive Applications, Paul R. Giblock

Masters Theses

There is an increasing gap between the rate at which data is generated by scientific and non-scientific fields and the rate at which data can be processed by available computing resources. In this paper, we introduce the fields of Bioinformatics and Cheminformatics; two fields where big data has become a problem due to continuing advances in the technologies that drives these fields: such as gene sequencing and small ligand exploration. We introduce high performance computing as a means to process this growing base of data in order to facilitate knowledge discovery. We enumerate goals of the project including reusability, efficiency, …

Visualization Techniques For Neuroscience-Inspired Dynamic Architectures, Margaret Grace Drouhard

Masters Theses

This work introduces visualization tools for Neuroscience-Inspired Dynamic Architecture (NIDA) networks and for the Dynamic Adaptive Neural Network Array (DANNA) hardware implementation of NIDA. A NIDA network is a novel type of artificial neural network that has performed well on control, anomaly detection, and classification tasks. We introduce a three dimensional visualization of software NIDA networks that represents network structure and simulates activity on networks. We present some of the analysis tasks for which the tool has been used, including the identification of useful substructures within NIDA networks through activity analysis and through the tracing of causality paths from events …

Data Analytics Of University Student Records, Mark Blaise Decotes

Masters Theses

Understanding the proper navigation of a college curriculum is a daunting task for students, faculty, and staff. Collegiate courses offer enough intellectual challenge without the unnecessary confusion caused by course scheduling issues. Administrative faculty who execute curriculum changes need both quantitative data and empirical evidence to support their notions about which courses are cornerstone. Students require clear understanding of paths through their courses and majors that give them the optimal chance of success. In this work, we re-envision the analysis of student records from several decades by opening up these datasets to new ways of interactivity. We represent curricula through …

Minimal-Density, Raid-6 Codes: An Approach For W = 9, Bryan Andrew Burke

Masters Theses

RAID-6 erasure codes provide vital data integrity in modern storage systems. There is a class of RAID-6 codes called “Minimal Density Codes,” which have desirable performance properties. These codes are parameterized by a “word size,” w, and constructions of these codes are known when w and w + 1 are prime numbers. However, there are obvious gaps for which there is no theory. An exhaustive search was used to fill in the important gap when w = 8, which is highly applicable to real-world systems, since it is a power of 2. This paper extends that approach to address the …

Mapping Spatial Thematic Accuracy Using Indicator Kriging, Maria I. Martinez

Masters Theses

Thematic maps derived from remote sensing imagery is increasingly being used in environmental and ecological modeling. Spatial information in these maps however is not free of error. Different methodologies such as error matrices are used to assess the accuracy of the spatial information. However, most of the methods commonly used for describing the accuracy assessment of thematic data fail to describe spatial differences of the accuracy across an area of interest. This thesis describes the use of indicator kriging as a geostatistical method for mapping the spatial accuracy of thematic maps. The method is illustrated by constructing accuracy maps for …

Automated Generation Of Simulink Models For Enumeration Hybrid Automata, David Aaron Heise

Masters Theses

An enumeration hybrid automaton has been shown in principle to be ready for automated transformation into a Simulink implementation. This paper describes a strategy for and a demonstration of automated construction. This is accomplished by designing a data model which represents EHA data and providing a mapping from EHA data points to Simulink blocks.

A Study Of Possible Optimizations For The Task Scheduler ‘Quark’ On The Shared Memory Architecture, Vijay Gopal Joshi

Masters Theses

Multicore processors are replacing most of the single core processors nowadays.

Current trends show that there will be increasing numbers of cores on a single chip in the coming future. However, programming multicore processors remains bug prone and less productive. Thus, making use of a runtime to schedule tasks on multicore processor hides most of the complexities of parallel programming to improve productivity. QUARK is one of the runtimes available for the multicore processors. This work looks at identifying and solving performance bottlenecks for QUARK on the shared memory architecture. The problem of finding bottlenecks is divided into two parts, …

Sequence Mining Based Debugging Of Wireless Sensor Networks, Kefa Lu

Masters Theses

Wireless Sensor Network (WSN) applications are prone to bugs and failures due to their typical characteristics, such as extensively distributed, heavily concurrent and resources restricted. It becomes critical to develop efficient debugging systems for WSN applications. A flexible and generic debugger for WSN applications is highly demanded. In this thesis, I proposed and developed a flexible and iterative WSN debugging system based on sequence analyzing and data mining techniques. At first, I developed vectorized Probabilistic Suffix Tree (vPST), a variable memory length model to extract and store sequential information from program runtime traces in compact suffix tree based vectors, based …

Tor Bridge Distribution Powered By Threshold Rsa, Jordan Hunter Deyton

Masters Theses

Since its inception, Tor has offered anonymity for internet users around the world. Tor now offers bridges to help users evade internet censorship, but the primary distribution schemes that provide bridges to users in need have come under attack. This thesis explores how threshold RSA can help strengthen Tor's infrastructure while also enabling more powerful bridge distribution schemes. We implement a basic threshold RSA signature system for the bridge authority and a reputation-based social network design for bridge distribution. Experimental results are obtained showing the possibility of quick responses to requests from honest users while maintaining both the secrecy and …