Innovative Projects Realized

Extending the level proximal subdifferential

Many real-world applications can be modeled as nonconvex optimization problems, such as the phase retrieval problem in medical imaging and matrix factorization problems in data analysis, just to name a few. Proximal-type algorithms serve as ideal candidates to resolve these nonconvex problems. Nevertheless, their convergence analysis remains challenging due to the lack of favorable properties and a uniform framework. Developing the novel extension of the level proximal subdifferential will enhance the mathematical foundation of proximal-type algorithms in the absence of convexity, anticipated to provide a uniform framework for the convergence analysis of these algorithms. As such, this project will bring new insights about the behavior of proximal-type algorithms for the nonconvex optimization problems, fostering better algorithmic solutions to computational challenges arisen from real-world scientific applications.

A Decision-Support System for Solar Photovoltaic Adoption in Sustainable Manufacturing based on the S5 Framework

This project is developing a decision-making tool to help small and medium-sized manufacturing companies in Canada and Mexico use solar power more effectively. This tool considers not just saving money and the environment, but also safety for workers and the community. By working together, researchers in Canada and Mexico will improve this tool and share their knowledge, benefiting both countries’ efforts to use clean energy.

Investigating the Synergistic Effect of Light-Inducing and Ultrasonication on the Decoration of Single Atoms on TiO2 Nanosheets for Enhanced Photocatalytic Hydrogen Generation

The project aims to advance clean energy technologies by addressing critical challenges in photocatalytic hydrogen generation using TiO2 nanomaterials. While TiO2 demonstrates strong photocatalytic abilities, harnessing solar energy for renewable energy production, it faces obstacles such as high recombination rates and slow kinetics for hydrogen evolution, necessitating the use of metal cocatalysts. Single-atom (SA) decorated catalysts show promise in enhancing efficacy and reducing system costs. This project focuses on two atomic-scale defect engineering techniques sonochemical and light-induced methods to decorate TiO2 nanosheets with single Pt atoms, increasing hydrogen generation rates. These methods create surface-exposed atomic-scale defects conducive to SA deposition, enabling efficient hydrogen production. The project’s outcomes are poised to propel sustainable energy technology, addressing challenges in clean energy production and promoting societal and environmental well-being. By sharing data and publishing papers resulting from this project, the University of Alberta and the University of Siegen can establish themselves as key players in the renewable energy industry, contributing to its development.

Les partenariats publics privés en sécurité au Québec : un état de la situation

Ce projet a pour objectif de dresser un panorama des partenariats public privé en sécurité (PPPS) au Québec. Alors que la gouvernance partenariale en sécurité se profile comme un moyen de plus en plus privilégié pour produire, distribuer et contrôler la sécurité dans nos sociétés contemporaines, peu d’études empiriques ont cherché à en évaluer le nombre et leur forme. Cette recherche, en partenariat avec le Bureau de la Sécurité Privée et la Banque Nationale du Canada, se donne comme ambition de combler ce vide, sur le territoire québécois. Elle servira à identifier les bonnes pratiques, ainsi que les limites et obstacles à ces nouvelles formes de collaboration, ce qui s’avérera utile aux décideurs politiques, aux organisations policières et aux acteurs de l’industrie de la sécurité privée.

Quantum sensor-based localization system for future urban air mobility

This project is investigation of conventional air vehicle localization using GPS and mobile networks and verification on that their weaknesses in the accurate and seamless localization can be supplemented or strengthened by future quantum sensor-based localization to discover a new blue ocean in the mobility industry.

Enhancing Liver MRI Diagnostics through Advanced 7T Imaging and Recurrent Inference Machine Technology

In this project, we aim to revolutionize liver MRI scans using advanced technology. Our focus is on utilizing a powerful 7T MRI scanner, which provides clearer and more detailed images than traditional MRI machines. We will be applying a cutting-edge technique called the Recurrent Inference Machine (RIM) to process these images. This method, developed by experts at UMC Amsterdam, uses sophisticated algorithms to reconstruct high-quality images quickly and efficiently. By doing so, we expect to achieve faster and more accurate liver scans, which is particularly crucial for diagnosing and treating various liver conditions. The outcome of this research will not only enhance MRI technology but also has the potential to significantly improve patient care, thanks to faster and more reliable diagnoses.

EPCOR / NAIT for a Summer COOP Student

This proposal is for the development of an exercise for catastrophic event that could significantly impact EPCOR’s operations and its ability to provide electricity, water or wastewater treatment services to over three million customers in Alberta. The findings from this project will significantly enhance EPCOR’s ability to plan for, respond to, continue operations and recover from such a large-scale event.

Prédiction de la demande chez les commerces de détail

Lila Solution offre une application de planification des achats (prévisions des ventes et gestion d’inventaire) afin de répondre aux défis majeurs auxquels font face les détaillants au quotidien : le manque d’outils efficaces pour la planification des achats et la gestion d’inventaire.

C’est dans ce contexte que notre solution a vu le jour. Notre plateforme révolutionnaire transforme la manière dont les détaillants abordent ces défis, remplaçant l’incertitude par la clarté. Grâce à nos outils, les détaillants prennent des décisions éclairées et maximisent leurs profits.

Lila Solution désire aller plus loin en développant un modèle de prédiction de la demande pour aider les commerçants à prédire ce qu’ils seront en mesure de vendre dans le futur en fonction de différents paramètres : nombre de produits, type de produits, inventaire en main au moment de la vente, tendance du marché, etc.

Development of Strategic Market Assessment Tools for Materials R&D

Chemia Discovery is a company that specializes in the research and development of new materials that can enhance the performance and sustainability of various technologies and enable emergent technologies. Chief among the materials explored at Chemia are those with applications for waste heat recovery and carbon capture. Chemia evaluates potential material R&D projects based on a comprehensive assessment of market potential, scientific viability, and technical considerations. Given the vast amount of information required for accurate decision-making, the R&D process for even a single material can be both time-intensive and resource-heavy. This project aims to enable Chemia to make assessments of the market potential for materials R&D, guiding its material R&D roadmap by mapping market data to technological and scientific data on specific classes of materials in target sectors. The main outcome of the project is the development of an easy-to-use and versatile tool designed to streamline the analysis of diverse market data, thereby facilitating strategic materials R&D planning.

Design a post-quantum cryptographic algorithm based on lattice for mitigating quantum computing threats.

Cryptographic algorithms are fundamental tools for securing digital information, with symmetric and asymmetric algorithms serving as the cornerstone of modern encryption techniques. Symmetric algorithms utilize a single key for both encryption and decryption, while asymmetric algorithms employ a pair of keys for these operations. Classical cryptographic algorithms, including RSA and AES, have been extensively utilized to ensure data security across various digital platforms.

However, the emergence of quantum computing poses a significant threat to the security provided by classical cryptographic algorithms. Quantum computers have the potential to exploit vulnerabilities in these algorithms using algorithms like Shor’s algorithm, which can efficiently factorize large prime number and solve discrete logarithmic problems. The advancement of quantum will imposes a threat to traditional cryptographic methods, prompting the need for exploring new solutions to maintain rigid security against evolving threats.

To address this vulnerability, we propose lattice-based cryptography as an advanced solution for post-quantum cryptography (PQC). Lattice-based cryptography offers resistance against quantum attacks due to its complex mathematical properties and hardness assumptions. The moto of our research is to successfully design and implement lattice-based cryptographic algorithm, providing a robust defense mechanism against potential quantum computing threats and contribute to the ongoing evolution of cybersecurity measures.

Exploring LLMs as a Foundation for Next-Generation Clinical Decision Support Systems  

The proposed project aims to develop an AI-powered Patient Assessment and Diagnostic (PAD) tool aimed at aiding in the diagnosis and management of chronic women’s health conditions. Leveraging Large Multi-Modal Models (LMMs), the project seeks to streamline the diagnostic process for hormonal health conditions, addressing the significant gap in timely healthcare access faced by millions globally. There are three main objectives including the design and development of an AI model with and without fine-tuning on a developed and curated women’s health dataset, validation using newly collected anonymized patient data, and optimization of output reports for clinicians, adhering to compliance standards.

Robust Portfolio Optimization

Financial investment is a crucial activity for both institutions and individuals. In particular, the construction of a portfolio of financial assets is at the core of financial investment decision making. Effective portfolio management enables many benefits such as better pensions and funds for education. The essence of portfolio construction is to balance expected returns (i.e. financial benefit) and risk (i.e. the possibility of financial loss). Quantitative methods play a crucial role in the generation of portfolios that optimizes the trade-off between risk and reward. However, most methods need estimates of important parameters such as future expected returns and other statistical quantities that go into the calculation of risk. A major limitation is that most models are not able to effectively incorporate that fact that these estimates change over time (e.g. due to unexpected turbulence in the markets) thus making previous estimates obsolete or misleading resulting in overly risky or underperforming portfolios. The research in this proposal aims to effectively mitigate parameter uncertainty by considering robust and dynamic models for portfolio optimization. As our partner organization Manulife Financial routinely constructs financial portfolios the research from this internship would be greatly beneficial for Manulife Financial.