Projets novateurs réalisés

Explorez des milliers de projets réussis issus de la collaboration entre organisations et talents postsecondaires.

30 508 projets complétés

2882
AB
5105
C.-B.
825
MB
681
NL
860
SK
9051
ON
9491
QC
97
PE
586
NB
1141
NS

Projets par catégorie

Federated Learning for Language Models

The existing approach to building Machine Learning models includes gathering all the data from customers in one place and then running the training procedure using it. However, there are no guarantees personal data won’t be leaked by the one training the model. So, the customers face a difficult decision of either allowing companies to gather their data to improve the products or restrict information sharing to ensure privacy. In this project, we eliminate the need to make such a decision by developing a system that would analyze the customers’ data without them sending it anywhere. We employ Federated Learning techniques, which allows training machine learning models on the data but not allowing the company to see the data. Our system targets Natural Language Processing applications: one of the most privacy-sensitive yet highly demanded areas of Machine Learning. To further foster customers’ trust, we use additional privacy methods to ensure no information can be inferred from the communication between customers’ devices and the company’s servers.

Voir la description complète du projet
Superviseur du corps professoral :

Aleksandar Nikolov

Étudiant :

Partenaire :

Microsoft Canada

Discipline :

Computer science

Secteur :

Information and cultural industries

Université :

University of Toronto

Programme :

Accelerate

Geomorphic changes in Russell Creek Experimental Watershed following a record rain on snow event

Movement of sediment through landslide activities dominates sediment transport in headwater channels, and exerts a strong control on downstream channel morphology and aquatic habitat. Landslide and debris flow activity in headwater channels is controlled in part by the intensity of precipitation events. At Russell Creek Experimental Watershed on northern Vancouver Island, high resolution, remotely sensed (LiDAR) topographic data is available from both before and after a record rain-on-snow hydroclimatic event. This unique dataset, coupled with a dense weather station network through the watershed, presents a unique opportunity to examine the geomorphic impacts of a record runoff event on debris flows, landslides, and downstream channel and alluvial fan morphology. This research will develop mitigation strategies for hazardous debris flow features which frequently close a highway on Vancouver Island, and will contribute to a growing body of research on the links between climate, hydrology, land use, geomorphology and aquatic habitat.

Voir la description complète du projet
Superviseur du corps professoral :

Marwan Hassan

Étudiant :

Partenaire :

Forestry BC (Prince George)

Discipline :

Sociology

Secteur :

Sustainability & the Environment

Université :

The University of British Columbia

Programme :

Accelerate

Diverging from the GPUs: The case for alternative architectures for training ML algorithms

Machine learning has ushered many breakthroughs in areas such as computer vision, natural language processing, speech recognition, and recommendation systems. The models used in these applications contain many parameters that need to be learned and training often requires massive amounts of computation. As such in recent years graphics processor units (GPUs) have seen wide adoption for the training of these large-scale models. While there has been a lot of work on GPUs, little focus has been given to other architectures for machine learning. As such, this project plans to explore a series of architectures that diverge from the beaten path of GPUs through modeling and simulation on state-of-the-art machine learning workloads. The study aims to determine the trade-offs of the architecture in terms of performance to help better design machine learning accelerators.

Voir la description complète du projet
Superviseur du corps professoral :

Maryam Mehri Dehnavi

Étudiant :

Partenaire :

AMD Canada

Discipline :

Computer science

Secteur :

Manufacturing; Professional, scientific and technical services

Université :

University of Toronto

Programme :

Accelerate

Essential Technologies for Autonomous Systems: Theory, Verification, and Applications

Autonomous Unmanned Vehicles (AUVs) are systems that are capable of maneuvering in the air, on the ground, above or under the water. There are a number of potential applications for AUVs in civilian, military, and security areas and AUVs have great potential benefits to Canada for numerous reasons. This proposed Mitacs-Alliance project aims to advance some essential technologies (cooperative control of multiple autonomous vehicles; high-precision navigation and positioning; optimal motion planning and decision-making) for autonomous unmanned vehicles in applications, such as autonomous transportation, self-driving cars, collaborative mapping, etc. The successful development of these technologies will further contribute to maintaining Canada’s status as a leading research country in the area of autonomous unmanned vehicles, and will, in addition, translate into monetary benefits for the Canadian companies who participate in these areas. This project is also important for developing expertise and talent within Canada needed for the next phases of AUVs research and development. The proposed research has the potential to enable new innovative products and services, based on autonomous unmanned vehicles, to be developed and produced in Canada. That would bring benefits of economic expansion and innovation savings.

Voir la description complète du projet
Superviseur du corps professoral :

Jinjun Shan

Étudiant :

Partenaire :

Quanser Inc;York University

Discipline :

Engineering

Secteur :

Education

Université :

York University

Programme :

Accelerate

SOTI Snap Analytics

Currently, data analytics services are not only expensive, but also often require days of delay for the data analytics report to be generated. The project focuses on making use of a user’s voice query as an input to create a data analytics response by aggregating data from a database. The successful implementation of the project suggests a novel way of providing data analytics services which significantly improves its accessibility. For example, if a user asks “What is my company’s revenue in the last two months?”, the model should be able to handle the query and automatically create an analytic response. To achieve such performance, the model needs to understand the semantics of query and knows how to identify and assemble desired data in a database.

Voir la description complète du projet
Superviseur du corps professoral :

Scott Sanner

Étudiant :

Partenaire :

SOTI Inc

Discipline :

Computer science

Secteur :

Information and cultural industries; Professional, scientific and technical services

Université :

University of Toronto

Programme :

Accelerate

Seasonal Change in Roosting Ecology in the Silver-haired bat (Lasionycteris noctivagans)

Silver-haired bats are common species of bat found in North America. They use cavities in trees and space under loose bark to roost, or rest and raise young. The silver-haired bat is thought to migrate south over the winter. Despite this, we have found them in parts of British Columbia during the winter, suggesting they may not migrate in these areas. Our work will help support a MSc student who will investigate how silver-haired bats are using trees in areas where they overwinter in British Columbia and compare with how they use trees in the summer. We will capture bats and track them to identify the tree roosts. Once located, we will record how long the bats stay at each tree, and tree features to identify patterns in tree use. Our partner, the Wildlife Conservation Society of Canada has identified this area of research as a high priority conservation action. The results from this project will be used to plan bat-friendly forestry practices by identifying trees that are suitable as winter and summer roosts for silver-haired bats.

Voir la description complète du projet
Superviseur du corps professoral :

Erin Baerwald

Étudiant :

Partenaire :

Wildlife Conservation Society Canada

Discipline :

Earth science

Secteur :

Professional, scientific and technical services

Université :

University of Northern British Columbia

Programme :

Accelerate

Évaluation du potentiel de croissance et de rendements d’une culture de tomates biologiques cultivées dans une serre climatisée à l’aide d’un système géothermique.

Deux serres expérimentales de tomates biologiques cultivées en plein sol sont utilisées dans le cadre de ce projet. La serre témoin est refroidie et déshumidifiée par ventilation naturelle tandis que la serre prototype utilise un système de géothermie. L’objectif du présent projet est de comparer et de quantifier les performances agronomiques des deux régies de culture. Les données de croissance des plants, le rendement, le calibre ainsi que certaines données physico-chimiques des fruits sont mesurés régulièrement pendant deux saisons de culture consécutives. Ces données serviront à vérifier l’hypothèse selon laquelle la production de tomates biologiques cultivées en plein sol peut être améliorée tant au niveau de la qualité que du rendement des fruits en utilisant un système de climatisation par géothermie. Si les résultats s’avèrent concluants, cette étude pourrait mener à l’instauration d’un système de refroidissement et de déshumidification par la géothermie aux Serres Jardins Nature.

Voir la description complète du projet
Superviseur du corps professoral :

Damien de Halleux

Étudiant :

Partenaire :

Les Serres Jardins Nature

Discipline :

Earth science

Secteur :

Agriculture

Université :

Université Laval

Programme :

Accelerate

VIRTAS – VIRtual Design and Testing of Aircraft Structure

The goal of the project is to develop an improved design process and tools for an aircraft structure based on digital simulation and artificial intelligence.
In order to minimize costs and improve the reliability of numerical simulations, we plan to reduce the number of physical tests, while optimizing the instrumentation. This will result in better quality data. An efficient method of processing this data and a decision support system in the event of measurements deviating from those simulated will also be developed.

Voir la description complète du projet
Superviseur du corps professoral :

Ilyass Tabiai;Martin Levesque

Étudiant :

Partenaire :

Bombardier Aerospace Inc (Dorval, QC)

Discipline :

Engineering

Secteur :

Manufacturing; Professional, scientific and technical services

Université :

École de technologie supérieure

Programme :

Accelerate

Dynamique de la pessière noire à peuplier faux-tremble de la plaine argileuse abitibienne : Une approche dendroécologique.

Ce projet vise à comprendre la dynamique forestière des peuplements d’épinette noire et de peuplier faux-tremble dans la pessière à mousse de la plaine argileuse abitibienne. Spécifiquement, il vise à comprendre si les peuplements se remplacent après feu sans changement d’espèces (dynamique cyclique, c’est-à-dire le peuplier est remplacé par du peuplier, et l’épinette noire par de l’épinette noire) ou si les forêts montrent une succession après feu du peuplier vers l’épinette noire avec le temps. Dans un secteur de 2000 km2 sélectionné dans la région écologique 6a2 (Plaine du lac Matagami), on procédera d’abord à caractériser la variabilité naturelle de la composition forestière à l’aide de données existantes et de travaux de terrain. Certains sites seront ensuite sélectionnés et échantillonnés afin de faire une reconstitution de leur développement à l’aide de la dendrochronologie.

Voir la description complète du projet
Superviseur du corps professoral :

Martin Simard

Étudiant :

Partenaire :

Norbord Inc (La Sarre, QC)

Discipline :

Earth science

Secteur :

Agriculture

Université :

Université Laval

Programme :

Accelerate

La participation des dirigeants au capital-actions a-t-elle une influence sur la performance de I’entreprise et sur son prix en bourse?

Le projet de recherche qui se tiendra chez Montrusco Bolton Investments tentera de faire la lumiere sur l’influence que les achats et ventes d’actions poses par la direction ont sur Ie prix en bourse de leurs entreprises. Cela tout particulierement au niveau des transactions d’inities, c’est-a-dire lorsqu’une personne a l’interieur de l’entreprise achete ou vend des portions de cette meme entreprise sur la bourse. Plusieurs etudes anterieures ont ete effectuees dans d’autres pays et sur d’autres variables. Par contre, cette recherche s’ avere unique et propose de regarder de nouvelles variables et de se baser sur un marche particulier, soit celui du Canada. En utilisant de nouvelles variables et un marche unique, on s’attend a de surprenantes conclusions qui valent largement la peine d’etre etudiees. Le sujet est d’ailleurs tres en vogue dans Ie domaine de la finance.

Voir la description complète du projet
Superviseur du corps professoral :

Alix Mandron

Étudiant :

Partenaire :

Placements Montrusco Bolton inc

Discipline :

Business

Secteur :

Finance and Insurance

Université :

HEC Montréal

Programme :

Accelerate

Estimation the cost of metal manufacturing

Cost estimation is a critical part of manufacturing. It has a significant role in determining the feasibility of decisions including design stage, planning and production methods. Especially, it is useful for subcontracting processes that GRAD4 conducts where the features of parts are submitted and bidding is the way of the deal between suppliers and buyers. We will investigate the historical cost and material data of parts and develop estimation methods utilizing statistical analysis and machine learning. The data includes features of the parts such as the operations needed, the geometry as well as the materials. The data particularly include three separate cost items: bending cost, laser cost, and material cost. The material cost has a deterministic feature that can be computed without variation. However, bending and laser costs has more complex nature which can be captured by pattern recognition applications. The final estimations would be in confidence interval form to enable the measurement of uncertainty and risk assessment.

Voir la description complète du projet
Superviseur du corps professoral :

Mehmet Gumus

Étudiant :

Partenaire :

GRAD4 Inc;Axya Inc.

Discipline :

Computer science

Secteur :

Professional, scientific and technical services

Université :

McGill University

Programme :

Accelerate

Effect of exposure to hydrocarbons on reproductive functions in select marine organisms

The partner organization, Oceans Ltd, offers diagnostic services to detect and monitor the effects of environmental contaminants on marine ecosystems, mainly using fish as biomonitoring tools. The company wishes to seize an opportunity for capacity building and diversification by exploring new bioindicator technologies based on the effects of hydrocarbons on the reproduction of marine invertebrates. The Mercier Laboratory will provide key expertise on the general biology and reproduction of marine invertebrates, which will be instrumental in exploring this avenue. The intern will use this knowledge to assist Oceans Ltd develop indicators of oil exposure that relate to reproductive functions in selected cold-water species. Specifically, the intern will be involved in carrying out the hydrocarbon exposures and subsequent biomarker analyses as part of his graduate (MSc) studies.

Voir la description complète du projet
Superviseur du corps professoral :

Annie Mercier

Étudiant :

Partenaire :

Oceans Ltd;Petroleum Research Newfoundland & Labrador

Discipline :

Life Sciences

Secteur :

Professional, scientific and technical services

Université :

Memorial University of Newfoundland

Programme :

Accelerate

Précédent
1...1,6011,6021,6031,6041,6051,6061,607...2,543
Suivant

Joignez-vous à un écosystème d’innovation florissant.

Abonnez-vous à notre infolettre dès maintenant!

S’inscrire