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 […]

Read More
Identifying Microaggressions Experienced by BIPOC Engineering Students across Higher Education in Ontario

Acknowledging that discrimination and prejudice of various sorts (e.g., verbal, behavioural, environmental) continue to exist in the education system, this research seeks to address how and why microaggressions against Black, Indigenous, and People of Colour (BIPOC), within engineering departments, show up among peers in classrooms, across interactions in lab environments, group-activities, and more. This research […]

Read More
Hydrologic regulating services of non-woody grassed and meadow urban landscapes

Due to urbanization in the GTA, many green areas are replaced by impervious infrastructure, preventing runoff infiltration and increasing flood risks. The Meadoway project proposed by the Toronto and Region Conservation Authority (TRCA) aims to restore meadow habitats with native plants along a hydro-corridor over 16 km across Scarborough, which can potentially improve the runoff […]

Read More
Personalized relevance content recommendation

PC Optimum was originally built as a rewards platform. As we are moving in the path to engage more meaningfully with our customers by way of personalized & interactive experiences (content), we are building a targeting engine as a solution to surface personalized content tailored for each individual customer. The problem we are solving then […]

Read More
AI-driven detection of anomalous supplier profiles at web scale

Information gathered from internet sources has high variance and different types of noise. This causes out-of-distribution problems with downstream ML modules such as category classification and keyword extraction. The extremely large size of internet-scale datasets requires a solution that is efficient and scalable. The objective of this project is to develop a start-of-the-art anomaly detection […]

Read More
Evolutionary Housing Strategies for Canadian Cities

This project examines the feasibility of low-cost evolutionary housing strategies to accommodate changing uses in Canada. It is a partnership between the University of Waterloo school of Architecture and architect John van Nostrand who has developed numerous prototypes for evolutionary housing in Canada and internationally. The research will begin with an analysis of existing prototypes […]

Read More
Homelessness Prevention at Covenant House Toronto

The proposed research project to be undertaken by the intern is concerning the use of preventative measures to combat youth homelessness and trafficking in Toronto through Covenant House Toronto. Youth homelessness and trafficking are on the rise and researching how prevention for this vulnerable population can be utilized has the potential to significantly reduce rates […]

Read More
Arts’ Civic and Social Impact Beyond Covid-19: Researchers in Residence

Four research interns will be supported by four arts funders, five universities and four arts and culture organizations, which will come together under the leadership of Mass Culture to see how the arts and culture sector impacts society today. Together, we will look at ways to strengthen community ties and foster shared social values through […]

Read More
Automatic Model Risk Management

Regulatory bodies, such as the Office of the Superintendent of Financial Institutions (OSFI) require that financial institutions properly assess and manage risk. Regulatory bodies aim to ensure the stability of financial markets, and the economy at large, and they do so by imposing guidelines on a firm’s risk management practices. Risk management models in general, […]

Read More