14277 research outputs found
Sort by
Efficient machine learning for Wi-Fi CSI-based human activity recognition using fast Monte Carlo based feature extraction
No full textWi-Fi sensing is gaining attention to recognize human activities without intruding
on people's privacy. Using existing wireless infrastructure, human activity can be
detected and classified without wearables or cameras. Once there is a movement
within a WI-Fi channel, the phase and amplitude are altered and are distinct for each
activity. The channel state information (CSI) contains detailed information about the
phase and amplitude within the channel and picks up subtle changes in the signal
caused by a person's movement through space. While CSI offers rich temporal and
spatial information, its high dimensionality poses significant challenges for real-time
processing, storage, and machine learning.
This thesis as presented in chapter 2, presents an efficient feature extraction
framework for CSI-based human activity recognition (HAR) that uses a fast Monte
Carlo-based algorithm grounded in the Frieze-Kannan-Vempala (FKV) low-rank
approximation technique, which is employed to extract a compact and discriminative
subset of features from the Doppler-transformed CSI data. The algorithm utilizes
length-squared sampling to identify high-energy rows and columns of the data matrix
and applies singular value decomposition (SVD) on a smaller submatrix to uncover
latent activity patterns. Rejection sampling subsequently selects features that align
with learned directions of maximum relevance, guided by label information. The
selected features significantly reduce the input dimensionality, enabling efficient training
of convolutional neural networks (CNNs), long-short-term-model (LSTM), CNN-LSTM,
and decision tree for activity classification. Experimental results demonstrate that
the proposed feature extraction method helps preserve classification accuracy while
substantially reducing computational complexity. This makes the approach suitable
for edge-device deployment and scalable HAR systems. This work advances low-cost,
privacy-preserving, and computationally efficient HAR by integrating signal processing,
randomized algorithms, and machine learning
Oxidative weathering and the genesis of continental red beds in the Huronian supergroup, Canada: new perspectives on continental paleoredox archives
No full textThe first appearance of oxidized paleosols and continental red beds in the Paleoproterozoic rock record is an important Fe-based proxy marker of newly oxidizing surface conditions accompanying Earth’s Great Oxidation/Oxygenation Event (GOE). However, there has been little attempt to bridge the independent records of these metasomatically overprinted rocks with a more unified understanding of Fe surface-cycling, from its source in weathering profiles to its sink in continental siliciclastic sedimentary deposits. This thesis contributes to this unification through 3 studies focused on rocks of the Gowganda and Lorrain formations of the ca. 2.45–2.22 Ga Huronian Supergroup (Ontario/Quebec, Canada) deposited during the GOE and across an interval of significant glacial-to-deglacial climate change. First, the metasomatic origin of alkali-element enrichments in paleosols is reviewed with literature data, and used to develop a new, integrated geochemical-petrographical approach to correction for excess K in paleosols formed on different rock types capturing redox-variable Fe weathering behavior. Second, a targeted field, mineralogical, and geochemical study revisiting the basement granite-hosted Ville-Marie paleosol and its capping polymictic breccia (Lorrain Formation) provides new evidence for lithology- and environment-dependent Fe behavior during in situ weathering and near-source transport, akin to modern redox-variable systems. Third, a broad-sampling, geochemical-mineralogical study of red beds across the Gowganda and Lorrain formations examines sedimentary Fe behavior across wider continental environments and into the shallow-marine realm, finding new robust evidence in hematite micromorphologies for syn- to early post-depositional Fe oxidation. The latter two studies better resolve Huronian red beds as a sedimentary product preferentially formed in: 1) catchments proximal to high-Fe mafic rocks; 2) source-distal environments only where fine-grained detritus could deposit; and 3) environments favoring in situ Fe oxidation during early sediment burial
Application of weather research and forecasting and coastal wave modelling to assess regional climate change impacts
No full textClimate change is accelerating, driving widespread environmental transformations,
with pronounced impacts on coastal regions. Rising sea levels, shifts in precipitation
patterns, changes in global wave climates, and increases in storm frequency and
intensity pose numerous challenges that must be considered for the safe and resilient
design of engineering structures. While many large-scale changes are well documented,
their local impacts remain less understood, particularly in regions with complex
geography. Newfoundland, directly exposed to the North Atlantic Ocean, faces
multiple climate-related hazards. Despite the recognized importance of these issues,
major knowledge gaps remain. This research addresses these gaps by integrating
high-resolution numerical modeling, programming, and data analysis, producing model
setups and datasets that provide a foundation for understanding localized climate
impacts and supporting future studies.
The study provides a comprehensive assessment of atmospheric, storm, and wave
dynamics, filling critical gaps in regional-scale understanding. The Weather Research
and Forecasting (WRF) model was applied to simulate near-surface atmospheric
conditions, capturing key spatial and temporal variability while identifying seasonal
biases. Historical and future climate projections were analyzed using the Weather
Research and Forecasting (WRF) model driven by General Circulation Model (GCM)
outputs, enabling the detection of climate change signals in key variables, including
maximum and minimum temperature, precipitation, and wind. Future projections
were assessed under the high-emission RCP8.5 scenario.
Automated Python-based storm tracking routines applied to WRF-derived sea-level
pressure data revealed changes in storm frequency and intensity. Empirical Orthogonal
Function (EOF) analysis was employed to identify dominant modes of atmospheric
variability and examine their evolution under future climate conditions. Changes in
wave climate were also evaluated using a cold-coupled wave modeling approach at
regional, local, and community scales, considering both mean conditions and extremes.
Multiple datasets were utilized throughout the study to establish the models, validate
simulations, and ensure the robustness of the results.
This research enhances the understanding of potential future climate scenarios by
quantifying projected changes in temperature, precipitation, wind, storm behavior,
and wave conditions, and by identifying areas vulnerable to extreme conditions. The
findings can inform future engineering designs and coastal protection strategies to
mitigate climate change impacts. Additionally, the developed models, methods, and
generated datasets can be utilized for future studies, expanding the potential impact
of this research
Sex, diet, and exercise-related differences on metabolic risk factors in C57BL/6J mice consuming diets with and without fructose
No full textHigh fructose consumption is associated with dysregulated carbohydrate and lipid metabolism. Dysregulated carbohydrate and lipid metabolism often precedes the development of type 2 diabetes and metabolic dysfunction-associated steatotic liver disease. Previous research suggests that dietary fructose alters metabolic outcomes in a sex-dependent manner. This study aimed to determine if voluntary wheel running, sex, and dietary fructose interact and impact body weight, liver weight, hepatic fat content, adipocyte morphology, energy expenditure, and normal glucose and lipid metabolism. 17 to 20-week-old male and female C57BL/6J mice (n=64), half of which had access to a running wheel, were fed high-fat diets containing 20% of total energy intake from fructose for 24 weeks. Comprehensive lab animal home cage monitoring systems were used to measure energy expenditure, RER, and lipid and glucose oxidation by indirect calorimetry. Before necropsy, mice received a 1:1 ratio of 0.5 g/kg glucose and fructose to assess postprandial fructose metabolism. When included with a high-fat diet, fructose did not exacerbate outcomes such as weight gain, serum LDL-cholesterol, hepatic steatosis, hepatic collagen deposition, and adipocyte size, and that sex had a greater impact on these outcomes. Previous exposure to fructose from diet significantly raised postprandial blood glucose levels. Sexual dimorphisms were seen in energy expenditure and lipid oxidation levels. Our data suggest that females have an enhanced ability to use dietary fat as fuel and that sex but not access to regular exercise impacts glycemia in mice consuming high levels of energy and fructose
Transdiagnostic factors of anxiety in math anxiety and math performance
No full textMathematics anxiety (MA) is a prevalent issue known to impact individuals of all ages around the world. It negatively impacts learning, performance, choice of program of study, and career choice. As a focus of research for decades, the impact of mathematics anxiety has been approached mainly from a pedagogical or cognitive perspective. While these areas have provided a wealth of knowledge, interestingly, less research has approached this anxiety from a clinical perspective, and little is known about how best to treat MA. This dissertation brings together current findings on MA as well as current findings from clinical psychology to provide a clinically relevant approach to the math-anxiety-math-performance relation to inform intervention.
Specifically, this dissertation focuses on the role of transdiagnostic factors of anxiety in MA. Beginning with an in-depth review of the current understanding of how MA negatively impacts performance and a review of four transdiagnostic factors: Emotion Regulation (ER), Emotional Distress Tolerance (EDT), Anxiety Sensitivity (AS), and Intolerance of Uncertainty (IU) in Chapter 1. Chapter 2, the first of two research papers, explores the moderating effects of the four transdiagnostic factors of anxiety in the math-anxiety-math-performance relation by assessing the influence of anxiety on performance by measuring two types of anxiety experienced during math tasks (i.e., state anxiety and traditional mathematics anxiety) and the impacts of these anxiety experiences on two types of math tasks (i.e., math facts, and math computation). Chapter 3 delves into developing a new measure of MA, the Subjective Experience of Math Anxiety (SEMA). This study explores the measure's psychometric properties, including internal consistency and validity. A shift away from the traditional measures of MA serves to provide a more clinically relevant way of measuring MA to inform treatment. Finally, Chapter 4 provides an in-depth review of the malleability of transdiagnostic
factors, and the treatments known to do so. It also serves to identify potential treatments that
could be adopted for use with mathematics anxiety
A "top-start" approach for reconstructing free-energy barriers using mean first-passage times
No full textRare events like crystal nucleation in mildly supercooled liquids occur infrequently
but proceed rapidly, resulting in the waiting time for their occurrence being much
longer than the timescale of the microscopic dynamics. Activated processes are analyzed
through free energy landscapes defined on one or more appropriate reaction
coordinates, where transitions between states require overcoming a barrier. In the
single reaction coordinate case, the mean first-passage time (MFPT) formalism [J.
Wedekind and D. Reguera, J. Phys. Chem. B, 112, 11060 (2008)] enables estimation
of transition rates and free energy landscapes using data from unbiased molecular
dynamics (MD) simulations, which is useful when the energy barrier is not too high.
For a sufficiently large barrier, spontaneous barrier crossings become rare and difficult
to observe within feasible simulation times, making the now-conventional MFPT approach
impractical. We extend the MFPT-based reconstruction method to overcome
the large barrier problem by starting MD trajectories near the transition state. To validate
our approach, we apply it to random walk models on a one-dimensional potential
energy landscape and compare our results with solutions from transfer matrices and
numerical integrals. Our findings show that the extended MFPT-based approach accurately
captures free energy landscapes and efficiently simulates activated processes
with our random walk model. We report on initial steps to test our method on crystal
nucleation in the Lennard-Jones liquid, comparing our results against the umbrella
sampling Monte Carlo method
Engineering education: pedagogy and practice within Newfoundland and Labrador's post-secondary engineering curricula
No full textThe purpose of this study is to examine the pedagogical practices of post-secondary engineering educators in Newfoundland and Labrador. Recognizing that instructional methods can play a critical role in student learning, the study explores current teaching strategies, the extent to which active learning principles are integrated into engineering instruction, and the ways digital technologies are employed to support the learning process.
Grounded in an interpretivist paradigm, the research employed a qualitative intrinsic single case study design with embedded units, capturing insights from educators and students across diverse engineering programs. The data collection instruments were comprised of an online questionnaire and semi-structured interviews designed to gather rich, contextual data from the study's participants. Subsequently, thematic analysis of the data revealed patterns associated with instructional practices, instructional preferences, challenges, and the use of digital technologies.
The findings revealed that while many educators value active learning and are increasingly experimenting with student-centered approaches, lecture-based instruction continues to dominate for a variety of reasons. Many students also expressed a preference for active learning strategies. In addition, the study identified a range of digital technologies supporting instruction, including specialized software and laboratory equipment.
This research contributes to provincial and national engineering education goals by showing areas of potential improvement and innovation in engineering pedagogy
Genotype/phenotype analysis of dilated and hypertrophic cardiomyopathy due to troponin complex variants in Newfoundland and Labrador families
No full textAbstract
Background/Objectives: Hypertrophic (HCM) and dilated cardiomyopathy (DCM) are inherited cardiomyopathies (prevalences 1:500 and 1:250 respectively). TNNI3 p.R162Q and TNNT2 p.ΔK210 are autosomal dominant pathogenic variants causing HCM and DCM respectively. Using large NL families segregating TNNI3 p.R162Q and TNNT2 p.ΔK210, this study will explore the penetrance, age of onset and severity of disease.
Methods: Six NL families have one or both variants with 79 affected (40 TNNI3 p.R162Q heterozygous, five TNNI3 p.R162Q homozygous, two TNNT2 p.ΔK210 heterozygotes, three double heterozygous) and 31 unaffected. All clinical cardiac tests were collected. Kaplan-Meier survivorship curves and cox regression analyses were performed using SPSS v.30.
Results: Twenty-three TNNI3 p.R162Q heterozygotes had clinical abnormalities, twelve had none. Median age for first clinical abnormality or SCD for affected individuals was 54yrs and 63yrs for unaffected (p=0.014, HR:2.684). Median age for affected males was 41yrs and 57yrs for females (p=0.019, HR:2.254). Two double heterozygotes show no structural signs of disease, the third had no cardiac concerns. TNNT2 p.ΔK210 heterozygotes have early, severe disease. TNNI3 p.R162Q homozygotes have highly variable phenotypes.
Conclusion: The cardiac phenotypes of TNNI3 p.R162Q heterozygotes and TNNI3 p.R162Q homozygotes are highly variable. Disease presents earlier in heterozygous males. TNNT2 p.ΔK210 phenotype is severe. Double heterozygotes have mild/no phenotype
Unveiling the seabird nesting history of Green Island, Witless Bay Ecological Reserve, NL using paleolimnology
No full textSeabirds are powerful biovectors that transfer nutrients across ecosystem boundaries;
from their marine foraging to their terrestrial breeding grounds. These nutrient inputs
accumulate in freshwater systems and can be traced through sediment analysis, offering a
long-term archive of colony presence and ecological impact. Reconstructing historical
seabird populations is essential for understanding long-term ecological change and informing
conservation strategies. This thesis reconstructs historical seabird dynamics on Green Island
(47°14'17" N, 52°46'49" W), a globally significant yet understudied seabird colony within
Newfoundland’s Witless Bay Ecological Reserve. Using a multi-proxy paleolimnological
approach, including δ¹⁵N isotopes, diatoms, pollen, chlorophyll a, and metal(loid)
concentrations, a dated 30 cm core (~170 years) extends population records beyond existing
data (~1940 CE). Congruent changes in all proxies suggest sustained ornithogenic enrichment
beginning in the late 1930s until present. However, δ¹⁵N peaks mid 20th century and may be
linked to shifts in prey composition associated with ocean warming. Aerial imagery between
1969 and 2023 further captured a decline in surface vegetation driven by excess nutrient input
from guano deposition. Distinguishing species-specific contributions to nutrient loading
remains challenging, particularly mixed species colonies. This study explored avian
papillomavirus (APV) as a potential molecular proxy by screening both sediments and
contemporary oropharyngeal and cloacal swabs. Although APV was not detected in
sediments, likely due to degradation or low abundance, viral DNA was recovered from both
Atlantic puffins (Fratercula arctica) and, for the first time in this region, from common
murres (Uria aalge). Minimal sequence divergence suggests a slowly evolving lineage and
possible interspecies transmission. While further research is needed to evaluate its
sedimentary potential, these findings underscore the promise of combining molecular and
paleolimnological approaches to improve species-level resolution in historical
reconstructions. By extending seabird population records beyond existing census data, this
research provides new insights into past seabird dynamics and their relationship with
environmental change
Trace elements and inclusions in metamorphic blue sapphire (Beluga deposit, Nunavut, Canada) Updated deposit model and implications for exploration and origin determination
No full textThis study examines the trace element geochemistry and mineral inclusions of a suite of sapphire corundum samples from the Beluga deposit, Lake Harbour Group, Baffin Island, Nunavut, to (1) refine the genetic model of the sapphires (with potential implications for exploration), (2) comprehend the observed colours and colour zoning, (3) assess the validity of geochemical source rock discrimination approaches in the case of the Beluga deposit, and (4) examine whether sapphires from Beluga can be reliably differentiated from sapphires from other deposits. Sixty (60) polished cross-sections were examined, taken from two parts of the Beluga pit from a total of 54 different crystals. Inclusions were characterized with SEM-EDS and EPMA. Minor and trace elements in corundum were measured using EPMA and LA-ICP-MS. The inclusions study shows that unaltered phlogopite, scapolite, and oligoclase were incorporated into growing sapphire corundum. These relationships are hidden in the host rock due to pervasive retrograde alteration, and the new data demonstrates corundum formation during M2 retrograde metamorphism, earlier than thought previously. Beluga sapphire is intensely oscillatory colour-zoned with alternating blue and colourless bands. The blue colour is caused by Fe²⁺-Ti⁴⁺ intervalence charge transfer (IVCT). Higher Fe and Ti concentrations correlate to some degree with a darker blue colour, while colourless zones generally contain less Fe+Ti, but with significant variability and overlap. Some colourless zones contain sufficient Fe and Ti, in theory, to create a blue colour. Possible reasons for the lack of colour in these zones may be incorrect Fe oxidation state (Fe³⁺) for Fe-Ti IVCT or an abundance of rutile nano-inclusions. While no exsolved rutile is observed with optical or electron microscopes, these zones can contain high Ti concentrations, which may suggest the presence of such inclusions. Beluga sapphire is geochemically identifiable as a metamorphic-type blue sapphire but is difficult to differentiate from Sri Lankan or Madagascan metamorphic blue sapphires using physical appearance, inclusions, and geochemistry. However, rutile silk, common in other blue sapphires, is not observed in Beluga sapphire. More investigation of inclusions, but on faceted gems using Raman spectrometry, is recommended