Graduate Theses & Dissertations

Exploring the Scalability of Deep Learning on GPU Clusters
In recent years, we have observed an unprecedented rise in popularity of AI-powered systems. They have become ubiquitous in modern life, being used by countless people every day. Many of these AI systems are powered, entirely or partially, by deep learning models. From language translation to image recognition, deep learning models are being used to build systems with unprecedented accuracy. The primary downside, is the significant time required to train the models. Fortunately, the time needed for training the models is reduced through the use of GPUs rather than CPUs. However, with model complexity ever increasing, training times even with GPUs are on the rise. One possible solution to ever-increasing training times is to use parallelization to enable the distributed training of models on GPU clusters. This thesis investigates how to utilise clusters of GPU-accelerated nodes to achieve the best scalability possible, thus minimising model training times. Author Keywords: Compute Canada, Deep Learning, Distributed Computing, Horovod, Parallel Computing, TensorFlow
Risk of Mortality for the Semipalmated Plover (Charadrius semipalmatus) Throughout Its Life Cycle
Three long-term mark and recapture/resight data sets of individually marked Semipalmated Plovers (Charadrius semipalmatus) were analyzed using Cormack-Jolly- Seber models. Data came from two breeding populations (Churchill, Manitoba, Canada, n=982, and Egg Island, Alaska, USA, n=84) and one overwintering population (Cumberland Island, Georgia, USA, n=62). For Alaska and Georgia, time-invariant models were best-supported, giving annual survival estimates of 0.67 (95%C.I.: 0.58- 0.76) and 0.59 (95%C.I.: 0.49-0.67) respectively. Data from Manitoba supported a timedependent model: survival estimates varied from 1.00 to 0.36, with lowest estimates from recent years, supporting observations of local population decline. Seasonal survival analysis of the Georgia population indicated lower mortality during winter (monthly Φoverwinter: 0.959, 95%CI: 0.871-0.988; for 6 month period Φoverwinter: 0.780 (0.440-0.929)) than during combined breeding and migratory periods (monthly ΦBreeding+Migration: 0.879 (0.825-0.918); for 8 month ΦBreeding+Migration: 0356 (0.215-0.504)). I recommend, based on high resight rates, continued monitoring of survival of wintering populations, to determine potential range-wide population declines. Keywords: survival, longevity, mortality, shorebird, overwinter, breeding, migration, life cycle Author Keywords: life cycle, longevity, mortality, non-breeding, shorebird, survival
Re-Living the Residential School Experience
The residential school legacy is one of the darkest chapters in Canadian history. From the mid-1850s to 1996, thousands of Aboriginal children were taken from their homelands and placed in residential schools. Taken against their will, many dreaded attending these schools. Some attended for as long as ten to fifteen years, only to be strangers in their own communities upon their return. In the past thirty years, survivors began disclosing the loneliness, confusion, fear, punishment and humiliation they suffered within these institutions, and also reported traumatic incidents of sexual, physical or emotional abuse. These childhood traumas still haunt them today. This dissertation examines the four compensation processes (Litigation, Alternative Dispute Resolution, Independent Assessment Process and the Common Experience Process) used by survivors to determine whether the compensation payments made to them assisted in reconciliation of their residential school experience. To complete an analysis of the processes, twenty-four residential school survivors from Ontario, Manitoba and British Columbia were interviewed about their experiences with one or more of the compensation processes. The study begins with a historical overview of the residential school legacy and continues with the residential school healing movement that initiated and finalized a negotiated settlement agreement for all living survivors. This dissertation provides a unique perspective to the residential school legacy by using a cultural framework, Anishinabe teachings and concepts to share the voices of residential school survivors. The pivotal Anishinabe teaching within this study comes from The Seventh Fire Prophecy. This prophecy states that: “If the New People will remain strong in their quest, the Waterdrum of the Midewiwin Lodge will again sound its voice.” In this dissertation the residential school survivors are the New People. As the dissertation unfolds the author utilizes various Anishinabe concepts to illustrate how the compensation processes failed to assist the New People to reconcile with their residential school experience. This study presents a medicine wheel understanding of reconciliation and the Residential School Legacy. It concludes with an important message to the second and third generation survivors to continue the reconciliatory efforts that the New People introduced. It is crucial that the children and grandchildren of the New People begin the reconciliation process not only for themselves but for the next seven generations. Author Keywords: Anishinabe, compensation, Indian residential schools, reconciliation, survivors
Effect of Listing a Stock on the S&P 500 Index on the Stock’s Volatility
This paper investigates the effect of listing a stock on the S&P 500 Index on the stock’s volatility, using various econometrics models: GARCH and EGARCH. The study mainly addresses three issues; firstly, it analyzes stock volatility in two sub-periods, secondly, it determines whether the announcement can account for the fluctuations in the price of the stock, and finally, it investigates the change in the stock’s variance. After isolating the effects of external and industry shock by using the returns on the S&P 500 Index as a proxy, the author finds evidence of structural change in the volatility of stocks after that stock is added to the index. Additionally, the existence of a dominant symmetric effect, which captures the response of volatility to news, indicate that following the onset of including the stock on the index, information flowing into the market increased. However, the rate at which old news is captured in price falls. The empirical evidence also suggests that on average a stocks variance falls and that the announcement to list a stock on the index has little effect on the stock’s price. Author Keywords: EGARCH, GARCH, S&P 500 Index, Symmetric Effect, Volatility

Search Our Digital Collections


Enabled Filters

  • (-) ≠ Reid
  • (-) ≠ English (Public Texts)
  • (-) = Williams
  • (-) ≠ Turmel
  • (-) ≠ Master of Arts

Filter Results


2010 - 2020
Specify date range: Show
Format: 2020/01/28

Author Last Name

Last Name (Other)