37200 research outputs found
Sort by
Death with Dignity: Queer Representation in Deathcare Systems
Get PDFThis thesis examines the ways that elderly members of the LGBT community lack representation within deathcare systems. Currently, the LGBT community is vastly underrepresented in research relating to the dying process and deathcare. This lack of representation creates barriers for LGBT individuals at the end of their lives and creates difficult end-of-life experiences. I utilized previous research that studied how LGBT populations interact with and are treated by a variety of death care-related systems, including health systems, legal systems, and therapeutic systems. This research highlighted the need for the LGBT community to be better represented within both research relating to these systems and within these systems directly. This thesis argues that without this representation, aging LGBT community members lack safety when dying, and that these systems need to be changed in order to achieve equitable deathcare
Beavers, Hydrology, and Wapato: A Baseline for Monitoring Franz Lake National Wildlife Refuge
Get PDFLocated in the lower Columbia River floodplain, Franz Lake National Wildlife Refuge is a unique landscape with a complex land use history. For thousands of years, Indigenous tribes lived on this land. In the early 1990s, U.S. Fish and Wildlife Service acquired the land, after it was identified as a mitigation site following the construction of the Bonneville Lock and Dam. Franz Lake Refuge was once known for its prevalent Wapato (Sagittaria latifolia) population, an emergent plant with edible tubers and an important food source for Indigenous people. With specific growth requirements and hydrologic conditions for germination and establishment, S. latifolia is susceptible to environmental changes. American beavers (Castor canadensis) can dramatically alter a landscape to suit their needs, such as digging channels, rerouting water, and through direct consumption of vegetation. To better understand the hydrology in Franz Lake Refuge, water elevation data was collected continuously, and historical lake surface water and beaver dams were evaluated using remote sensing. Wetland vegetation transect surveys were performed and used for an image classification of the Franz Lake Refuge floodplain. The results suggest that there are many factors influencing the hydrology and wetland plant community within Franz Lake Refuge. The Columbia River has a significant influence on Franz Lake Refuge, especially during large flooding events; however, the effects from beaver dam building activity on Franz Lake hydrology were inconclusive. In addition, reed canary grass (Phalaris arundinacea) has become a dominant species and is likely encouraging on S. latifolia habitat. This project was designed to create a baseline for future monitoring and anticipated habitat restoration
Modeling Injury Severity of Truck-Involved Crashes Under COVID-19 Regulatory Changes
No full textAt the start of the pandemic the Federal Motor Carrier Safety Administration issued declaration No. 2020-002 which included a long list of goods given temporary emergency relief (deregulation) from national statutes 49 CFR Parts 390-399 of the Federal Motor Carrier Safety Regulations that oversee trucking. In particular, this granted emergency relief from Title 49 CFR § 395.3: maximum driving time for motor carriers or their drivers, regarded as hours-of-service (HOS), and subsequently further expanded to include short-haul trips, adverse driving conditions, break requirements, and sleeper berths. Regulations which play a crucial role in establishing the safe operation and management of motor carriers, their drivers, and the public.
The consequences for these suspensions of regulations from a safety standpoint are not fully understood. The research presented here endeavors to provide insight into these motor carrier regulation relaxations from a safety perspective by assessing crash proportions and modeling injury severity. Oregon crash data from 2019 through 2021 (before, during, and after COVID) was used for analysis. Proportions tests indicate statistically significant differences between time periods for various crash-related factors, while the injury severity models suggest contributing injury severity factors were also different among time periods
Designing for Deployable, Secure, and Generic Machine Learning Systems
Get PDFMachine learning systems have catalyzed numerous image-centric applications owing to the significant achievements of machine learning algorithms and models. While these systems have showcased the efficacy of machine learning models, certain challenges persist, such as machine learning system design and security vulnerabilities inherent in deep neural networks. Moreover, the deployment of deep neural network models remains a significant hurdle. This dissertation introduces a multimedia prototyping framework tailored for visual analytical applications, improving the reusability of video analysis software tools with minimal performance overhead. Furthermore, we present novel image-processing techniques designed to bolster the robustness of deep neural networks and propose an innovative compression technique to address deployment challenges.
First, we propose a new software prototyping framework called Video as Text (vText) that analyzes and manipulates the video data as trivial as we handle text data in most Unix and Linux systems to tackle the reusability issue in the existing video analysis tools. The vText paradigm seeks to mimic such programs. We demonstrate the design and implementation of vText linking video codecs with computer vision and image processing algorithms, and the performance evaluation shows that the vText framework achieves comparable running time and is easily used for prototyping visual analytical programs.
Second, to reduce the vulnerability of deep neural networks against adversaries, we propose three color-reduction image processing approaches, which are Gaussian smoothing plus PNM color reduction (GPCR), Gaussian smoothing plus K-means (GK-means), and fast GK-means to make deep convolutional neural networks more robust to adversarial perturbation. We evaluate the approaches on a subset of the ImageNet dataset. Our evaluation reveals that our GK-means-based algorithms have the best top-1 classification accuracy.
The final contribution of the dissertation is introducing a novel deep neural network compression framework on class specialization problems to address the limited utilization of deep neural network-based functionalities. We propose a novel knowledge distillation framework with two proposed losses, Renormalized Knowledge Distillation (RKD) and Intra-Class Variance (ICV), to render computationally efficient, specialized neural network models. Our quantitatively empirical evaluation demonstrates that our proposed framework achieves significant classification accuracy improvements for the tasks where the number of subclasses or instances in datasets is relatively small
Three Essays on Communicative Planning: From the Perspective of East Asians
Get PDFCommunicative planning aims to strengthen the inclusiveness of planning by shifting the paradigm from the perception that the public is considered an object affected by the planning process to an active participant. In modern society, where people\u27s values are more diverse, communicative planning can be an alternative way to reflect the opinions of people with diverse identities for decision-making. At the same time, however, it is criticized for its difficulty in universal application, particularly in non-Western countries. Scholars argue that since communicative planning theory is biased toward Anglo-Americans\u27 context, it requires a specific contextual condition.
In this background, this dissertation seeks to understand communicative planning in the East Asian context through various approaches. Specifically, the dissertation consists of three essays on communicative planning in the East Asian context from both theoretical and empirical perspectives. Considering the cultural uniqueness of non-white racial groups, this dissertation pays attention to East Asians, who have been marginalized in both communicative planning and, more broadly, the field of urban planning.
The first essay conducts a systematic literature review to look at the situated communicative planning theory and practices in three East Asian countries: China, South Korea, and Japan. The second essay provides an empirical examination of how the communicative planning process operates in East Asian countries, using South Korea as an example. This essay particularly focuses on the role of facework, which plays a crucial role in human relationships in East Asia. Using qualitative content analysis with official reports, news articles, and video recordings of the community forum, this essay reveals that facework complicates the attainment of the ideal conditions of communicative planning, but it facilitates a cooperative attitude among participants. The final essay explores the racial disparity of neighborhood associations from the perspective of East Asians in Portland, Oregon. Using the mixed method approach, this essay shows that East Asian Portlanders are underrepresented in neighborhood associations, even though Asians are the second most populous racial group in Portland. Additionally, this essay identifies a significant barrier to initial participation: the mismatch between the characteristics of neighborhood associations and those of East Asian Portlanders
Undergraduate Students of Color Raising Children and Persisting in Higher Education
Get PDFMillions of undergraduate students have been identified as parents across the United States. Of those millions, a majority have been identified as undergraduate students of Color who are pregnant, parenting, frequently underrepresented, and often not equitably supported toward degree completion. The purpose of the qualitative single-site case study was to learn what undergraduate students of Color who are parenting have experienced, in terms of support for their continued enrollment, while earning baccalaureate degrees at Portland State University. Through an asset-based approach, this study elaborated on three key findings which include the persistent aspirations of undergraduate Students of Color (USPs of Color), despite the challenges they have faced, their resistance strategies, and their experiences with resources and resourcefulness. Recommendations offered in this study for future practice can be adopted by postsecondary education leaders, and community service leaders alike, to develop resources, improve current efforts, or amend policies and practices to equitably serve USPs of Color toward degree attainment
A Hierarchical Decision Model for Evaluating the Strategy Readiness of Quantitative Machine Learning/Data Science-Driven Investment Strategies
Get PDFBig data and computational technologies are increasingly important worldwide in asset and investment management. Many investment management firms are adopting these data science methods and technologies to improve performance across all investment processes. Researchers actively use these methods to develop more effective systematic investment strategies and produce more reliable outcomes less vulnerable to human decision-making biases. However, the success of such a strategy depends heavily on the scientific rigor applied throughout the process. Best practices involve understanding how to make better decisions in the research design process. A good question is whether we can make better decisions in developing quantitative strategies. Therefore, the decisions made in the research process are crucial to developing successful quantitative strategies. Additionally, as this field is inherently multidisciplinary, it requires a system thinking approach to consider multiple perspectives to provide a clearer understanding of the strategies often referred to as black boxes.
Therefore, the main objective of this research is to develop a multi-criteria assessment framework and scoring decision support system to evaluate quantitative investment strategies that apply machine learning and data science techniques in their research and development. Subject matter experts will assess all framework perspectives from a systematic literature review to approve their reliability. The perspectives consist of economic and financial foundations, data perspective, features perspective, modeling perspective, and performance perspective. The research methodology applied is the Hierarchical Decision Model (aka HDM) to provide a 360-degree view of the quantitative investment strategy and improve and generalize the concept to other asset classes and regions. Finally, this research helps investment researchers and professionals to focus on research process decisions in generating more hypotheses and developing financial theories to be tested empirically rather than cherry-picking investment strategies based on historical simulations
Is This My Place? Contributing Factors to Community College Students\u27 Longitudinal Sense of Belonging and the Connection of Sense of Belonging to Student Success
Get PDFCommunity colleges serve a large percentage of historically under-represented populations including students of color, low-income, and first-generation students. Unfortunately, less than half of students who begin at a community college return the following Fall and only about one-fourth of students eventually graduate. Previous research indicates that students\u27 sense of belonging may be a key factor in retention and completion rates. However, large-scale datasets have yet to comprehensively explore students\u27 sense of belonging in the community college setting. To address that issue, the purpose of this quantitative exploratory study was to examine individual and institutional variables associated with a measure of students\u27 institutional sense of belonging in their first and third years of school. Additionally, the research assessed how sense of belonging was related to measures of student academic success. The data are a subset of community college students who participated in the nationally representative NCES 2012-17 Beginning Postsecondary Students Longitudinal Study (BPS) (n=6,700).
Analyses revealed that student perceptions of faculty and peer interactions, satisfaction with academic and social experience, and academic confidence were most related to student sense of belonging. Differences were found between contributors to first- and third-year sense of belonging and between student demographic groups. Sense of belonging was found to increase the likelihood of students being retained at an institution and of completing a degree or certificate. By focusing on factors most related to community college students\u27 sense of belonging, institutions can leverage resources to support student retention and degree completion
Towards a New Discourse on Success in Alternative Education
Get PDFAlthough researchers agree that Alternative Education (AE) within the United States is an essential set of schools and programs that do things beyond traditional education, they do not agree on the purpose for these efforts. To understand how researchers can connect between existing perspectives and consider new ways that they can discuss success in the future, I interviewed 16 students and 15 staff from three different AE schools within the same state. Through thematic analysis, I found students and staff to describe success as an amalgamation of individual and common conceptions, requiring individual effort and support, a commitment to daily action as well as a wider, longer-term vision. AE students and staff also shared extensively about the role constructive relationships played in their pursuit of success. AE students and staff help broaden the ways researchers and policy makers can look at success in the future. They also provide glimmerings of what relationships based on wider conceptions of power can look like, something with increasingly wide implications when we broaden our considerations to relationships between not just students and staff, but also parents, the district, the local community, and society at large
Embodying Whole Community Preparedness: A Case Study of Student-Facing Preparedness Initiatives at Institutions of Higher Education in the State of Oregon
Get PDFOver the last two decades, there has been a dramatic increase in natural and manmade disasters. While no community is immune to the risks posed by these large-scale emergencies, colleges, and universities, which are home to numerous vulnerable populations, could be counted among those most at risk. In their current state of overall community preparedness, campuses are not equipped to weather impending dangers; consequently, the impact on students could be disastrous. This study examined preparedness practices at Oregon colleges and universities to better understand how institutions of higher education in the state prepare their students for large-scale emergencies, what challenges they face in implementing preparedness interventions, and what best practices have helped them overcome these challenges. Using a qualitative case study research design, an inventory of preparedness interventions was developed through the distribution of a state-wide qualitative survey, document-mining and elite interviews with preparedness managers at seven colleges and universities in Oregon. These interventions, in conjunction with data collected regarding challenges and best practices, were used to inform the development of a tool based on the Extended Parallel Process Model to aid practitioners in developing preparedness education interventions