Physical Sciences and Mathematics Commons^™

An Application Of Deep Learning Models To Automate Food Waste Classification, Alejandro Zachary Espinoza

Dissertations and Theses

Food wastage is a problem that affects all demographics and regions of the world. Each year, approximately one-third of food produced for human consumption is thrown away. In an effort to track and reduce food waste in the commercial sector, some companies utilize third party devices which collect data to analyze individual contributions to the global problem. These devices track the type of food wasted (such as vegetables, fruit, boneless chicken, pasta) along with the weight. Some devices also allow the user to leave the food in a kitchen container while it is weighed, so the container weight must also …

Local Radiance, Scott Peter Britell

Dissertations and Theses

Recent years have seen a proliferation of web applications based on content management systems (CMS). Using a CMS, non-technical content authors are able to define custom content types to support their needs. These content type names and the attribute names in each content type are typically domain-specific and meaningful to the content authors. The ability of a CMS to support a multitude of content types allows for endless creation and customization but also leads to a large amount of heterogeneity within a single application. While this meaningful heterogeneity is beneficial, it introduces the problem of how to write reusable functionality …

A Computational Model For Recovery From Brain Injury, Wayne Wakeland

Systems Science Friday Noon Seminar Series

A computational simulation model calculates recovery trajectories following traumatic brain injury (TBI). Prior publications include a multi-scale framework for studying concussion and a systems-level causal loop diagram (CLD) and discussion of feedback processes. The scope of the computational model goes beyond concussion to include all severities of TBI. A set of first order ordinary differential equations and their associated parameters determines recovery trajectories. While highly speculative, the model serves to demonstrate the potential utility of computational models in this context. Much more research will be needed to create a properly supported research model that could be used for clinical trial …

Fractals As Basis For Design And Critique, John Charles Driscoll

Dissertations and Theses

The design profession is responding to the complex systems represented by architecture and planning by increasingly incorporating the power of computer technology into the design process. This represents a paradigm shift, and requires that designers rise to the challenge of both embracing modern technologies to perform increasingly sophisticated tasks without compromising their objective to create meaningful and environmentally sensitive architecture. This dissertation investigated computer-based fractal tools applied within a traditional architectural charette towards a design process with the potential to address the complex issues architects and planners face today. We developed and presented an algorithm that draws heavily from fractal …

Correct-By-Construction Typechecking With Scope Graphs, Katherine Imhoff Casamento

Dissertations and Theses

Dependently-typed languages are well-known for the ability to enforce program invariants through type signatures, and previous work establishes the effectiveness of this style of program verification in the implementation of type-safe interpreters for a wide class of languages with a variety of interesting scoping semantics, offering an account of dynamic semantics. This thesis covers the complementary topic of static semantics, in the form of a pattern for constructing verified typechecking procedures in a dependently-typed setting. Implementations are given for simply-typed lambda calculus and a small procedural language as well as a module system with unrestricted cyclic module dependency semantics that …

Sensory Relevance Models, Walt Woods

Dissertations and Theses

This dissertation concerns methods for improving the reliability and quality of explanations for decisions based on Neural Networks (NNs). NNs are increasingly part of state-of-the-art solutions for a broad range of fields, including biomedical, logistics, user-recommendation engines, defense, and self-driving vehicles. While NNs form the backbone of these solutions, they are often viewed as "black box" solutions, meaning the only output offered is a final decision, with no insight into how or why that particular decision was made. For high-stakes fields, such as biomedical, where lives are at risk, it is often more important to be able to explain a …

Versatile Binary-Level Concolic Testing, Bo Chen

Dissertations and Theses

Computing systems are experiencing an explosive growth, both in complexities and diversities, ushered in by the proliferation of cloud computing, mobile computing, and Internet of Things. This growth has also exposed the consequences of unsafe, insecure, and unreliable computing systems. These all point to the great needs of sophisticated system validation techniques. Recent advances in research on symbolic execution has shown great promises for automated software analysis, e.g., generating test cases, finding bugs, and detecting security vulnerabilities. However, symbolic execution is mostly adopted to analyze user applications, while modern computing systems in practice consist of many components shipped by various …

A Secure Anti-Counterfeiting System Using Near Field Communication, Public Key Cryptography, Blockchain, And Bayesian Games, Naif Saeed Alzahrani

Dissertations and Theses

Counterfeit products, especially in the pharmaceutical sector, have plagued the international community for decades. To combat this problem, many anti-counterfeiting approaches have been proposed. They use either Radio Frequency Identification (RFID) or Near Field Communication (NFC) physical tags affixed to the products. Current anti-counterfeiting approaches detect two counterfeiting attacks: (1) modifications to a product's tag details, such as changing the expiration date; and (2) cloning of a genuine product's details to reuse on counterfeit products. In addition, these anti-counterfeiting approaches track-and-trace the physical locations of products as the products flow through supply chains.

Existing approaches suffer from two main drawbacks. …

Design And Experimental Evaluation Of Deepmarket: An Edge Computing Marketplace With Distributed Tensorflow Execution Capability, Soyoung Kim

Dissertations and Theses

There is a rise in demand among machine learning researchers for powerful computational resources to train complex machine learning models, e.g., deep learning models. In order to train these models in a reasonable amount of time, the training is often distributed among multiple machines; yet paying for such machines (either through renting them on cloud data centers or building a local infrastructure) is costly. DeepMarket attempts to reduce these costs by creating a marketplace that integrates multiple computational resources over a distributed TensorFlow framework. Instead of requiring users to rent expensive GPU/CPUs from a third-party cloud provider, DeepMarket allows users …

A Resource Constrained Shortest Paths Approach To Reducing Personal Pollution Exposure, Elling Payne

REU Final Reports

As wildfires surge in frequency and impact in the Pacific Northwest, in tandem with increasingly traffic-choked roads, personal exposure to harmful airborne pollutants is a rising concern. Particularly at risk are school-age children, especially those living in disadvantaged communities near major motorways and industrial centers. Many of these children must walk to school, and the choice of route can effect exposure. Route-planning applications and frameworks utilizing computational shortest paths methods have been proposed which consider personal exposure with reasonable success, but few have focused on pollution exposure, and all have been limited in scalability or geographic scope. This paper addresses …

Context-Aware Wi-Fi Infrastructure-Based Indoor Positioning Systems, Huy Phuong Tran

Dissertations and Theses

Large enterprises are often interested in tracking objects and people within buildings to improve resource allocation and occupant experience. Infrastructure-based indoor positioning systems (IIPS) can provide this service at low-cost by leveraging already deployed Wi-Fi infrastructure. Typically, IIPS perform localization and tracking of devices by measuring only Wi-Fi signals at wireless access points and do not rely on inertial sensor data at mobile devices (e.g., smartphones), which would require explicit user consent and sensing capabilities of the devices.

Despite these advantages, building an economically viable cost-effective IIPS that can accurately and simultaneously track many devices over very large buildings is …

A Computational Model For Recovery From Traumatic Brain Injury, Wayne Wakeland, Erin S. Kenzie

Systems Science Faculty Publications and Presentations

A computational simulation model calculates estimated recovery trajectories following traumatic brain injury (TBI). Prior publications include a multi-scale conceptual framework for studying concussion, a systems-level causal loop diagram (CLD) and an analysis of key feedback processes. A set of first order ordinary differential equations and their associated parameters determines recovery trajectories. The model contains 15 state variables, 73 auxiliary variables, and 50 parameters describing TBI pathology in an aggregate fashion at the cellular, network, cognitive and social levels. There are 1200 feedback loops, which give rise to a variety of behavior modes, many of which are highly nonlinear. Exogenous parameters …

Event Trend Aggregation Under Rich Event Matching Semantics, Olga Poppe, Chuan Lei, Elke A. Rundensteiner, David Maier

Computer Science Faculty Publications and Presentations

Streaming applications from cluster monitoring to algorithmic trading deploy Kleene queries to detect and aggregate event trends. Rich event matching semantics determine how to compose events into trends. The expressive power of stateof- the-art streaming systems remains limited since they do not support many of these semantics. Worse yet, they suffer from long delays and high memory costs because they maintain aggregates at a fine granularity. To overcome these limitations, our Coarse-Grained Event Trend Aggregation (Cogra) approach supports a rich variety of event matching semantics within one system. Better yet, Cogra incrementally maintains aggregates at the coarsest granularity possible for …

Crumpled And Abraded Encryption: Implementation And Provably Secure Construction, Scott Sherlock Griffy

Dissertations and Theses

Abraded and crumpled encryption allows communication software such as messaging platforms to ensure privacy for their users while still allowing for some investigation by law enforcement. Crumpled encryption ensures that each decryption is costly and prevents law enforcement from performing mass decryption of messages. Abrasion ensures that only large organizations like law enforcement are able to access any messages. The current abrasion construction uses public key parameters such as prime numbers which makes the abrasion scheme difficult to analyze and allows possible backdoors. In this thesis, we introduce a new abrasion construction which uses hash functions to avoid the problems …

Exploring And Expanding The One-Pixel Attack, Umairullah Khan, Walt Woods, Christof Teuscher

Student Research Symposium

In machine learning research, adversarial examples are normal inputs to a classifier that have been specifically perturbed to cause the model to misclassify the input. These perturbations rarely affect the human readability of an input, even though the model’s output is drastically different. Recent work has demonstrated that image-classifying deep neural networks (DNNs) can be reliably fooled with the modification of a single pixel in the input image, without knowledge of a DNN’s internal parameters. This “one-pixel attack” utilizes an iterative evolutionary optimizer known as differential evolution (DE) to find the most effective pixel to perturb, via the evaluation of …

Development Of A Design Guideline For Pile Foundations Subjected To Liquefaction-Induced Lateral Spreading, Milad Souri, Arash Khosravifar

Student Research Symposium

Past earthquakes confirmed that seismically induced kinematic loads from soil lateral spreading and inertial loads from structure can cause severe damages to pile foundations. The research questions are:

• How to combine inertial and kinematic loads in design of pile foundations in liquefied soil?
• How the combination of inertia and kinematics changes with depth?
• How this combination is affected by long-duration earthquakes?
• How this combination affects inelastic demands in piles?

Localizing Little Landmarks With Transfer Learning, Sharad Kumar

Dissertations and Theses

Locating a small object in an image -- like a mouse on a computer desk or the door handle of a car -- is an important computer vision problem to solve because in many real life situations a small object may be the first thing that gets operated upon in the image scene. While a significant amount of artificial intelligence and machine learning research has focused on localizing prominent objects in an image, the area of small object detection has remained less explored. In my research I explore the possibility of using context information to localize small objects in an …

High-Speed Video From Asynchronous Camera Array, Si Lu

Computer Science Faculty Publications and Presentations

This paper presents a method for capturing high-speed video using an asynchronous camera array. Our method sequentially fires each sensor in a camera array with a small time offset and assembles captured frames into a high-speed video according to the time stamps. The resulting video, however, suffers from parallax jittering caused by the viewpoint difference among sensors in the camera array. To address this problem, we develop a dedicated novel view synthesis algorithm that transforms the video frames as if they were captured by a single reference sensor. Specifically, for any frame from a non-reference sensor, we find the two …

Good Similar Patches For Image Denoising, Si Lu

Computer Science Faculty Publications and Presentations

Patch-based denoising algorithms like BM3D have achieved outstanding performance. An important idea for the success of these methods is to exploit the recurrence of similar patches in an input image to estimate the underlying image structures. However, in these algorithms, the similar patches used for denoising are obtained via Nearest Neighbour Search (NNS) and are sometimes not optimal. First, due to the existence of noise, NNS can select similar patches with similar noise patterns to the reference patch. Second, the unreliable noisy pixels in digital images can bring a bias to the patch searching process and result in a loss …

Efficient And Scalable Event Tracing, Rupika Dikkala

University Honors Theses

In this work, I demonstrate that a time series database can be utilized to store Open Trace Format 2 (OTF2) file metadata for common trace events efficiently and scalably. This paper examines the efficacy of storing event trace data in a time series database, and investigates associated performance overhead compared to the state of the art method using OTF2 trace files. The sample traces used in this project are generated from a parallel hydrodynamic modeling code, Lulesh, developed at Lawrence Livermore National Laboratory. In my approach, I first cache common event trace metadata in InfluxDB, a contemporary time series database. …

Artificial Intelligence Hits The Barrier Of Meaning, Melanie Mitchell

Computer Science Faculty Publications and Presentations

Today’s AI systems sorely lack the essence of human intelligence: Understanding the situations we experience, being able to grasp their meaning. The lack of humanlike understanding in machines is underscored by recent studies demonstrating lack of robustness of state-of-the-art deep-learning systems. Deeper networks and larger datasets alone are not likely to unlock AI’s “barrier of meaning”; instead the field will need to embrace its original roots as an interdisciplinary science of intelligence.

Spectral Clustering For Electrical Phase Identification Using Advanced Metering Infrastructure Voltage Time Series, Logan Blakely

Dissertations and Theses

The increasing demand for and prevalence of distributed energy resources (DER) such as solar power, electric vehicles, and energy storage, present a unique set of challenges for integration into a legacy power grid, and accurate models of the low-voltage distribution systems are critical for accurate simulations of DER. Accurate labeling of the phase connections for each customer in a utility model is one area of grid topology that is known to have errors and has implications for the safety, efficiency, and hosting capacity of a distribution system. This research presents a methodology for the phase identification of customers solely using …

The Complexities Of Open Data, Hector Dominguez

Systems Science Friday Noon Seminar Series

Hector Dominguez is the current Open Data Coordinator at the City of Portland, and there are several lessons learned and strategies developed in the several months of work in this position. Hector will share some challenges on creating trusted and reliable data and information services, as well as the opportunities to work with Urban Data to resolve city challenges and to support achieving the City's goals in the coming years.

In this talk, Hector will share how modeling and defining the right metrics are not the only factors to implementing a citywide program, but rather, how ethics, communications and strategy …

Knowing Without Knowing: Real-Time Usage Identification Of Computer Systems, Leila Mohammed Hawana

Dissertations and Theses

Contemporary computers attempt to understand a user's actions and preferences in order to make decisions that better serve the user. In pursuit of this goal, computers can make observations that range from simple pattern recognition to listening in on conversations without the device being intentionally active. While these developments are incredibly useful for customization, the inherent security risks involving personal data are not always worth it. This thesis attempts to tackle one issue in this domain, computer usage identification, and presents a solution that identifies high-level usage of a system at any given moment without looking into any personal data. …

Contingent Requirements For Artifical Intelligent Systems Development, Gary Langford, Herman Migliore

Engineering and Technology Management Faculty Publications and Presentations

A substantial portion of project failures are due to poorly defined requirements before enough is known about pragmatic end-item product capability, technology maturity, or development strategy. Process models either start with requirements or are weakly structured to elicit and derive actual stakeholder needs and to establish incontrovertible requirements. Existing process models are used acceptably for systems but are wholly inadequate for system and system of systems requirements that involve interactions with humans at a personal level. Problems with products and services are notable when artificial intelligent systems are put into use. Rather than establishing a technology baseline then working up …

Strategic Technology Planning In Product-Service Systems With Embedded Customer Experience Requirements, Soheil Zarrin, Tugrul Daim

Engineering and Technology Management Faculty Publications and Presentations

The undeniable impact of Artificial Intelligence and Internet of things on value proposition and offerings of firms, drive many strategic initiatives in organizations to design solutions which integrate products and services. Since designing Product-Service Systems inherently introduce high level of complexity and adding artificial intelligence requirements as one of the influential factors overcomplicate the long-term planning processes, the strategic planners seek for effective tools to enable them to manage the level of complexity as well as empowering them to communicate the outcomes with the whole organization. In order to achieve this purpose, Technology Roadmaps provides a structured and flexible means …

The Changing Moral Mirror Of Society: From Human To Artifical Intelligent Systems, Gary Langford, Teresa Langford

Engineering and Technology Management Faculty Publications and Presentations

Management of technology and its development carry along the responsibility and consequences for interactions between Human and Artificial Intelligent Systems (AIS). In spite of all good intentions, the effects and repercussions of conflicts between Human and the systems built with intent to assist Human may be proceeding along the path that will recognize a dismal mistake in judgment. Dreadful and intolerable impositions on Human behavior may arise regardless of how AIS is designed. That is not to say progress should cease, but rather to make the case that intensely determined efforts need to delve into the uses and implications of …

Domain Process Model Overcome Limitations Of Engineering Models For Developing Artificial Intelligent Systems, Gary O. Langford, John Green, Daniel P. Burns, Alexander Keller, Dean C. Schmidt

Engineering and Technology Management Faculty Publications and Presentations

The integrated set of prognostic domains (ISPD) of technology presented here provides a normative means to construct a wholly new process model for guiding Technology Management of Artificial Intelligent Systems (AIS). Seventeen domains represent all-inclusive stakeholder perspectives that encapsulate lifecycle analyses, evaluations, feasibilities, and tradeoffs with the domain contexts. Following Systems Model-Based thinking (SMBT), a postulated focal point interaction is the entry condition from which each domain is considered and thereafter traversed. Domains are interactive with each other through concurrent, iterative, recursive, and non-recursive processes. This interactive work continues until the completion milestones of each domain are satisfied. Techniques such …

High-Speed Video From Asynchronous Camera Array (Poster), Si Lu

Computer Science Faculty Publications and Presentations

Poster presented at: 2019 IEEE Winter Conference on Applications of Computer Vision (WACV)

Good Similar Patches For Image Denoising (Poster), Si Lu

Computer Science Faculty Publications and Presentations

Patch-based denoising algorithms like BM3D have achieved outstanding performance. An important idea for the success of these methods is to exploit the recurrence of similar patches in an input image to estimate the underlying image structures....