Search UNB Scholar
About UNB Scholar
UNB Scholar is an institutional repository initiative of UNB Libraries intended to collect, preserve, showcase, and promote the open access scholarly output of the UNB community. Use UNB Scholar to explore specific collections, or search all content in the repository. Material submitted to the repository will also be freely discoverable online through Google and other major search engines.
UNB Libraries welcomes submissions to UNB Scholar. Find out more about depositing your publications!
Communities in UNB Scholar
Select a community to browse its collections.
Recent Submissions
Exploiting temporal dynamics to improve the robustness of continuous myoelectric control
(University of New Brunswick, 2025-01) Tallam Puranam Raghu, Shriram; Scheme, Erik J.; MacIsaac, Dawn T.
Myoelectric control based on Surface electromyography pattern recognition (sEMG-PR) offers intuitive and dexterous control of powered prostheses for people with limb differences. However, conventional sEMG-PR systems often struggle with transitions between movements, impacting online usability. In this thesis, we investigated these transition-specific challenges and proposed novel approaches to enhance the performance and user experience of sEMG-PR systems.
We first established a comprehensive framework for evaluating classifier performance during transitions, incorporating transition-specific metrics and continuous dynamic datasets. This framework represents an improvement over conventional evaluation methods, which often focus primarily on steady-state performance and neglect transitions. Our analysis, utilizing this enhanced framework, revealed that classifiers, even with similar steady-state performance, can differ substantially in their ability to handle transitions. This finding underscores the limitations of conventional evaluation methods.
Next, we systematically investigated various error-mitigation strategies, including existing and novel post-processing techniques. While some techniques showed promise, particularly those based on rejection, our findings suggest that relying solely on post-hoc error correction may not be sufficient to address the challenges of transitions effectively.
Finally, we explored incorporating continuous dynamic data, inclusive of transitions, into the training process. Our results demonstrated the advantages of leveraging Long Short-Term Memory (LSTM) networks, which can effectively capture the dynamic nature of transitions. Furthermore, we pioneered the use of self-supervised learning for sEMG-PR, and demonstrated its effectiveness in learning meaningful and robust representations from unlabeled continuous dynamic data, leading to enhanced performance both offline and online.
Our findings underscore the crucial role of temporal information, dynamic training data, and appropriate model selection, particularly temporal models like LSTMs, in achieving robust and reliable sEMG-PR based myoelectric control. The proposed approaches have the potential to significantly enhance the usability and effectiveness of these systems, paving the way for more intuitive and user-friendly prosthetic devices.
Learning dynamic regimes of event-based substructures in EEG data using Graph Kernel Koopman Embedding
(University of New Brunswick, 2025-01) Nagawara Muralinath, Rashmi; Mahanti, Prabhat K.
Understanding brain activity requires analyzing EEG data, which is challenging due to the high noise levels, non-linearity, non-stationarity, and individual variability. This thesis introduces a novel methodology using Graph Kernel Koopman Embedding (GKKE) methodology by representing time-evolving brain connectivity as low-dimensional, meta-stable regimes. The study focuses on two critical applications: detecting epileptic seizures (CHB-MIT dataset) and assessing cognitive workload (Cognitive Mental Workload dataset).
This research attempts to classify cognitive and neurological states using various combinations of connectivity measures, graph kernels, and classifiers. The results demonstrate that the method has a good classification accuracy of above 85% for both datasets, thus demonstrating its potential to identify intricate patterns. The suggested method involves preprocessing the raw EEG data through which the connectivity matrix is obtained by calculating correlation coefficients and generating gram matrices. Next, we use kernel PCA to simplify the graph features by reducing their dimensions. After that, we test how well they work with machine learning classifiers.
Genetic variation in adventitious rooting, seed germination, and berry phenolic content of black elderberry (Sambucus canadensis) in New Brunswick
(University of New Brunswick, 2025-01) Germaine, Tanya Rae; Sacobie, Charles; Smith, Ron
Black elderberry (Sambucus canadensis), a North American shrub valued for its ecological and medicinal properties, remains underexplored compared to its European counterpart, Sambucus nigra. This study investigates genetic variation in seed germination, adventitious rooting, and phenolic content (chlorogenic acid and rutin) among wild S. canadensis populations in New Brunswick, Canada. Ten populations from diverse biogeographic zones were sampled. Germination success varied significantly (59%–78%), with coastal populations germinating faster. Phenolic concentrations ranged widely (chlorogenic acid: 487–1825 ng; rutin: 884–2404 ng), showing strong correlation (β = 0.735, p < 0.001). Root development showed limited site variability and no correlation with plant size. Results highlight substantial genetic and phenotypic diversity, underscoring the species’ potential for ecological restoration, sustainable agriculture, and bioactive compound production. This research informs population selection for adaptability and enhanced bioactive compound production.
Explainable decision-making framework concealed in graph statistical models of chemistry
(University of New Brunswick, 2025-01) El-Samman, Amer; De Baerdemacker, Stijn
Recent advancements in probabilistic models in chemistry have unlocked ground-breaking potential, yet these innovations come with heightened caution. Decisions made by techniques, such as neural network models, are seldom fully understood, even by developers themselves, making it difficult to integrate these models into an established scientific discourse. Nevertheless, their use remains widespread and likely to increase, as they generate predictions that rival or surpass traditional chemical models in efficiency. This great potential combined with a lack of explainability has placed these models under increasing scrutiny, leading to the field of explainable artificial intelligence.
This thesis investigates graph probabilistic models of chemistry, particularly graph neural nets, to develop an explanatory framework of decision-making that can be quantitatively blueprinted and replicated. We probe the cryptic high-dimensional nature of the feature space of these models, compacting their dimensions to elucidate a decision-making framework based on the molecular substructures of chemistry. We then demonstrate that the decision-making framework of these models is organized around chemical formula language/syntax from which the hidden framework can be replicated, while also providing a novel way of exploring reactions. Finally, we show the completeness of these models by transferring their capabilities to solve a wide range of chemical problems, from predicting pKa values and NMR data to modeling electron density and solubility.
The role of landscape features on the distribution of freshwater mussels in the Lower Wolastoq, New Brunswick
(University of New Brunswick, 2025-01) Cusack, Sarah; Gray, Michelle
Freshwater mussels provide habitat, structural stream bed support, nutrient cycling and act as an indication of ecosystem health. There are eight species in the lower Wolastoq: Alewife Floater (Utterbackiana implicata), Eastern Floater (Pyganodon cataracta), Triangle Floater (Alasmidonta undulata), Eastern Elliptio (Elliptio complanata), Eastern Lampmussel (Lampsilis radiata), Tidewater Mucket (Atlanticoncha ochracea), Eastern Pearlshell (Margaritifera margaritifera), and Yellow Lampmussel (Lampsilis cariosa). This study assessed the distribution of available habitat for in relation to the landscape controls producing differences in regional energetic gradients. Species-specific habitat distribution models were validated using snorkel surveys at 34 sites. This study determined that landscape features can be employed to identify suitable habitat across the Wolastoq. Suitable habitat characteristics did not always guarantee rare species presence but identified areas for future conservation work. Lower-gradient habitats supported higher species richness, while higher gradient habitats hosted fewer specialist species. Sediment characterization revealed species-specific preferences for depositional or erosional habitat.