Theses, Dissertations, and other Required Graduate Degree Essays

Receive updates for this collection

This collection contains digitized SFU theses except for those theses submitted within the last 12 months. If you cannot find the thesis you are looking for please search Recently Submitted Theses as it may be a recently submitted thesis and thus not yet available in Summit.

WiFi-based activity recognition with deep learning

Author: 
Date created: 
2020-05-26
Abstract: 

Human activity recognition is drawing escalating attention in recent years in both academia and industry due to the potentials in bracing such a broad range of Internet of Things (IoT) applications as health diagnosis, human-machine interactions, safety surveillance, and so on. Among many forms of sensing technologies, e.g., using cameras, wearable sensors, and RFIDs, WiFi-based activity recognition is of particular interest given its ubiquity, low cost, device-free experience, and low dependence. Generally, people's motions will affect the reflected WiFi signals and incur specific radio patterns. Through profiling these specific patterns, we are able to recognize the original activities. Many existing works have reported relatively good activity recognition performance in dedicated scenarios; yet their performance degrades much in the practical complex applications with various impact factors, such as the co-channel interference, spatial diversity, and diverse environments, making existing WiFi-based solutions far from being satisfactory. In this thesis, we aim to address the existing key challenges and develop accurate, reliable, and adaptive WiFi-based human activity recognition systems. We argue that the integration of advanced deep learning techniques into the activity recognition will bring new opportunities towards our goal. Along this end, we first propose CSAR, a channel selective activity recognition framework that conquers the channel quality problem by active channel hopping and channel combination. We then develop WiSDAR, which constructs multiple separated antenna pairs and obtains features from multiple spatial dimensions to solve the spatial diversity problem. We at last investigate the activity recognition in a more compact in-car scenario and present WiCAR, a WiFi-based in-car activity recognition framework that leverages domain adaptation to remove the environment-specific information in the received signals while retaining the activity-related features for adaptive recognition. We have conducted extensive evaluations and the performance results further demonstrate the superiority of our frameworks over the state-of-the-art solutions.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Jiangchuan Liu
Department: 
Applied Sciences: School of Computing Science
Thesis type: 
(Thesis) Ph.D.

Examining and modelling students’ selection of course modality

Author: 
Date created: 
2020-05-08
Abstract: 

Despite online courses’ growing popularity, the factors that shape undergraduates’ choice of course modality are still poorly understood. This study explores the relations between a wide range of factors and students’ modality selection, in a context where both modalities — face-to-face and online — were made available. Undergraduates from a Canadian University enrolled in face-to-face (N = 335) and online courses (N = 315) completed a questionnaire assessing personal factors, course attributes, goal orientation and learning strategies. Data were subject to descriptive and inferential statistical analysis, and two logistic regressions were performed to model students’ enrolment and preference. Analysis revealed that the groups differed significantly in twelve variables. For example, number of previous online courses and enjoyment of online courses were significantly higher for online students. Logistic regression analysis extended these findings, indicating ten significant predictors for online enrolment, among them higher number of previous online courses and higher work-avoidance goals.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Kevin O'Neill
Department: 
Education: Faculty of Education
Thesis type: 
(Thesis) M.A.

Use of functional correlation tensors for fMRI monitoring of neuroplasticity during motor learning

Author: 
Date created: 
2020-05-11
Abstract: 

The development of Functional Correlation Tensors (FCT) is driving novel investigations into whole-brain functional magnetic resonance (fMRI) signal synchronicity. FCTs are mathematically analogous to the established structural diffusion modality. Unlike conventional fMRI analysis, FCTs examine functional signal independently of hemodynamic response assumptions. In this work participants trained on a motor task for two weeks with fMRI and diffusion scans collected at baseline and endpoint. Using only baseline data, a significant correlation was detected for the fractional anisotropy of the diffusion data with the signal synchronicity anisotropy of the FCT data. Previous work on this data detected white matter (WM) neuroplasticity in motor regions between baseline and endpoint. As FCT is sensitive to WM function, it was hypothesized that WM neuroplasticity could be further detected. Significant increases in signal synchronicity were detected in areas of motor task planning and execution. This represents the first instance of this novel methodology for identifying neuroplasticity.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Ryan D'Arcy
Department: 
Applied Sciences: School of Engineering Science
Thesis type: 
(Thesis) M.A.Sc.

The price is not right: Reducing mobile telecommunications bills in Canada

Author: 
Date created: 
2020-04-21
Abstract: 

Over the past decade, the Canadian mobile telecommunications service market has grown at an exceptional rate, ultimately overtaking landlines as the dominant form of telecommunication. Mobile telecommunications has therefore become an essential service in the lives of most Canadians. However Canadian consumers continue to face some of the highest mobile service prices in the developed world. This study seeks to identify the principal causes of continued high mobile telecommunications prices in Canada, and to suggest potential policy solutions to lower service costs for consumers. This study involves a combination of expert interviews, secondary research, and case study analysis of the Australian, U.S. and Saskatchewan mobile markets. This study identifies high levels of market concentration as a principal cause, and recommends that mandated MVNO access be implemented to address high prices in the Canadian mobile service market.

Document type: 
Graduating extended essay / Research project
File(s): 
Senior supervisor: 
Joshua Gordon
Department: 
Arts & Social Sciences: School of Public Policy
Thesis type: 
(Project) M.P.P.

Kwaskastahsowin (“Put things to right”): Case studies in twentieth-century Indigenous women’s writing, editing, and publishing in Canada

Author: 
Peer reviewed: 
No, item is not peer reviewed.
Date created: 
2020-04-14
Abstract: 

In the updated and restored 2019 edition of Halfbreed, Métis writer Maria Campbell introduces the Cree word, kwaskastahsowin, to describe what it means to seek conciliation or to “put things to right.” By focusing on what it means to “put things to right” in the context of twentieth-century publishing in Canada, this dissertation considers the ways that Indigenous women writers working within the Canadian publishing industry have been negatively impacted by intersecting issues of colonialism, race, and gender. My project explores Campbell’s definition of kwaskastahsowin in relation to two key twentieth-century works of Indigenous women’s writing in Canada: E. Pauline Johnson’s Legends of Vancouver (1911) and Maria Campbell’s Halfbreed (1973). Using a decolonial case-study approach that combines archival/digital methods with Indigenous editorial principles and protocols, my project focuses on the literary and storytelling contributions of three Indigenous women authors – E. Pauline Johnson (Mohawk), Mary Agnes Capilano (Sḵwx̱wú7mesh), and Maria Campbell (Métis) – and situates their authorial contributions and publishing experiences within the larger context of twentieth-century Canadian publishing. My project uses archival research to interrogate the publishing contexts and histories of these texts, and reveals the extent to which colonial issues of voice and editorial intervention have shaped the published works. Furthermore, by approaching the various editions of these two texts with a focus on their shifting “paratexts” (Genette) or critical frameworks over time, I draw attention to the lasting impacts of such editorial interventions (as evidenced by the two excised and recently-recovered pages from Campbell’s Halfbreed, detailing her sexual assault by members of the RCMP, and through evidence of Johnson’s preferred title for the Legends of Vancouver collection). My examination of the publishing histories of these two key Indigenous texts reveals the urgency with which other works of twentieth-century Indigenous literature must be re-examined, and simultaneously calls for a necessary reenvisioning of the Indigenous literary paratext – one that takes into account Indigenous editorial principles and protocols.

Document type: 
Thesis
Senior supervisor: 
Sophie McCall
Department: 
Arts & Social Sciences: Department of English
Thesis type: 
(Thesis) Ph.D.

Angry Raven and Friends: Three new stories for Hul’q’umi’num’ language learners

Author: 
Date created: 
2020-04-28
Abstract: 

Storytelling is an important tool for sharing knowledge and language across generations. Stories teach us about our way of life and our perspectives on how to be as First Nations peoples. In this project, I share three new stories that I have created inspired by real-life experiences—the importance of singing in the Quw’utsun’ culture, the fixation of the younger generation on video games, and the cultural activities of our people as witnessed by a young sasquatch. Each story has an important life lesson that is presented through humour. Together with elders Delores Louie and Ruby Peter, I have brought these stories to life in Hul’q’umi’num’, a Coast Salish language of British Columbia. These stories are designed to engage the younger generation and inspire them to gain fluency in the Hul’q’umi’num’ language.

Document type: 
Graduating extended essay / Research project
File(s): 
Senior supervisor: 
Donna Gerdts
Department: 
Arts & Social Sciences: Department of Linguistics
Thesis type: 
(Project) M.A.

Identification and management of wasabi pathogens in British Columbia

Author: 
Date created: 
2020-03-27
Abstract: 

Wasabi (Wasabia japonica) plants in British Columbia are grown in moist conditions ideal for pathogens, and therefore, are prone to various diseases. Over 3 years, seven wasabi greenhouses were surveyed for pathogens. Prevalence and severity of diseases were documented. Pathogenic species including Phoma wasabiae (Leptosphaeria biglobosa), Botrytis cinerea, and Erysiphe cruciferarum were found in multiple greenhouses. A new disease of wasabi with symptoms of vascular blackening and wilt was discovered. Using morphological and molecular techniques, the causal organism was identified as Verticillium isaacii. Powdery mildew of wasabi caused by E. cruciferarum was prevalent in half the greenhouses surveyed. In order to evaluate management options for powdery mildew, 4 commercially available products, Actinovate®, Cueva®, Rhapsody®, and Regalia® were applied biweekly onto greenhouse plants. Both Cueva® and Regalia® significantly reduced the progression of powdery mildew on wasabi plants.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Zamir Punja
Department: 
Science: Biological Sciences Department
Thesis type: 
(Thesis) M.Sc.

The relationship between fracture morphology and bone biomechanics: A study of changes occurring in juvenile porcine ribs over the early postmortem interval

Date created: 
2020-03-06
Abstract: 

Forensic anthropologists are often faced with the challenge of determining fracture timing based on bone features, usually discussing it in terms of a “fresh” versus “dry” bone response. Yet it is still unclear how long into the postmortem period bone can retain its fresh characteristics, particularly juvenile rib bone. Juvenile porcine ribs were used to examine 1) changes in the biomechanical properties of bone in response to localized load, and 2) changes in the morphology of the resulting fractures over the early postmortem interval (PMI). Two macroenvironments (subaerial and burial) were recreated in a greenhouse. Samples were placed on the surface of soil filled containers and distinct samples were buried in the same containers, with a total of 16 containers being studied over 12 months. Samples were collected weekly for the first 4 weeks, the subsequent two were collected 2-weeks apart, and the remaining 4-weeks apart. Individual ribs 8-11 were selected from the subaerial (n=146) and the burial (n=134) environments and fractured experimentally. Six biomechanical parameters were collected from each test and median values were obtained for each sample. Each fracture was then examined for eight morphology characteristics and frequencies were calculated for each sample. In the subaerial samples, multiple regression analysis showed that displacement at peak force, displacement at failure and failure stiffness were significantly associated with the PMI. Type of fracture, presence of plastic deformation and presence of cortical peeling were also significantly associated with the PMI. In the buried samples, multiple regression analysis showed no significant association between bone biomechanics and the PMI, and only a moderate association was found between the PMI and fracture morphology, specifically in the type of fracture, fracture surface and presence fiber pull-out. Although a transition from “fresh” to “drier” bone was apparent in the subaerial samples, a persistence of typical “fresh” bone response over the year-long PMI was evident in the buried environment. Accurate timing assessment of juvenile rib fractures is thus likely to be compromised from the analysis of bone features alone and further investigations are necessary for more confident and accurate rib trauma analysis, specifically when involving child remains.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Hugo Cardoso
Department: 
Environment: Department of Archaeology
Thesis type: 
(Thesis) Ph.D.

Ecological consequences of flow regulation by Run-of-River hydropower on salmonids

Author: 
Date created: 
2020-04-02
Abstract: 

Streams are dynamic, disturbance-driven ecosystems, where flow plays a dominant role structuring biological communities. Anthropogenic activities on streams change natural patterns of flow and disturbance, which in turn alters the conditions to which resident fishes are adapted, and their survival and fitness. Run-of-river (RoR) hydropower projects are an example of an anthropogenic activity that may alter stream ecosystems by temporarily diverting a proportion of stream flow to produce electricity. RoR hydropower projects have increased considerably in number and importance in the last three decades in both British Columbia and worldwide. Although there is a perception that RoR hydropower has minimal effects on stream ecosystems due to the small physical footprint of projects, we know surprisingly little about the impacts of RoR hydropower on fish populations. In this thesis, I use a combination of published research, empirical data, and models to evaluate a range of hypotheses regarding how RoR hydropower may affect fish populations, concentrating on salmonid species whose freshwater habitats often overlap with RoR projects. In Chapter 2, I synthesize the impact pathways by which RoR hydropower may influence salmonid populations, inferred from studies of reservoir-storage hydropower and salmonid ecology. In Chapter 3, I use empirical data to quantify increases in water temperature due to RoR flow diversion and explore the possible consequences for resident fish growth with bioenergetics models. In Chapter 4, I evaluate how stranding from anthropogenic flow fluctuations affects the long-term population dynamics of coho salmon (Oncorhynchus kisutch) in RoR-regulated rivers using a matrix model integrating both the strength and timing of freshwater density dependence. Finally, in Chapter 5, I quantify the high level of uncertainty in how much compensation habitat is required to offset chronic mortality incurred by multiple life-stages of coho salmon. The global emergence of RoR hydropower projects emphasizes the importance of understanding their effects on aquatic ecosystems. Overall, our capacity to protect and restore threatened salmonid populations rests upon our ability to not only better understand the pathways of impacts, but also the effectiveness of both natural (density dependence) and human (habitat compensation) interventions that can be used to offset anthropogenic mortality.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Wendy J. Palen
Department: 
Science: Biological Sciences Department
Thesis type: 
(Thesis) Ph.D.

3D convolutional neural networks for Alzheimer’s disease classification

Author: 
Date created: 
2020-01-21
Abstract: 

Dementia of the Alzheimer’s type (DAT) is a neurodegenerative disease characterized by abnormal brain metabolism and structural brain atrophy. These functional and structural changes can be observed in images acquired using 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) and structural magnetic resonance imaging (MRI). Traditional machine learning framework for DAT classification often involves time-consuming segmentation of brain images as part of the feature extraction process, while deep neural networks can learn DAT-related patterns directly from brain images to generate DAT probability scores. In this thesis, we design 3D convolutional neural networks (CNN) for two applications: classification and segmentation. We design classification networks for single modality use and perform comprehensive evaluation by measuring the performance of our networks on images along the entire DAT spectrum. To support traditional DAT classification framework, we design a fast and accurate segmentation pipeline. We propose a hemisphere-based approach where we train networks to localize and segment hemispheres.

Document type: 
Thesis
File(s): 
Senior supervisor: 
Faisal Beg
Department: 
Applied Sciences: School of Engineering Science
Thesis type: 
(Thesis) M.A.Sc.