This project will look at improving the quality of software by using AI to determine if a defect / issue exists and if so where it exists for easier fixes. This research is innovative and this domain is not proven. The student will explore new techniques for a highly relevant issue in industry. The industry partner will gain insight and knowledge into how improvements can be made ultimately resulting in faster time to market and potential cost savings.
In Ontario, 95% of its paved roads in the province are paved with asphalt or asphalt surface treated. Due to severe weather conditions, the lifetime of asphalt roads is relatively short and regular maintenance is required. The annual cost of maintenance is estimated to be around 2.14 billion dollars. This project aims to explore nanotechnology to increase the lifetime of asphalt roads by exploring nanomaterials with anti-oxidation activities. Liposomes will be used as a model system for the initial studies and asphalt doped liposomes and finally asphalt will be tested.
The field of artificial intelligence is traditionally divided into two broad paradigms. On one hand there are symbolic, formal, procedural, deterministic, and/or rule-based methods that often rely on a set of atomic elements and rules operating on those elements. Sometimes they are as complex as a comprehensive reasoning system. It is relatively effortless to provide a few manual instructions to these systems, however, these instructions (i.e., rules) are labor-intensive and become unfeasibly time-consuming as the complexity of the system grows beyond a certain point.
Risk aggregation is omnipresent in insurance applications. A recent example, borrowed from the modern regulatory accords, is the determination of the aggregate economic capital and its consequent allocation to risk drivers. A more traditional illustration of the importance of risk aggregation in insurance is the celebrated collective risk theory that dates back to the early years of the 20th century. This project will assist Sun Life Financial to build and implement an efficient quantitative framework to approximate the aggregate risk of its portfolio.
Theft or loss of sensitive data is a growing concern for companies who may suffer losses of consumer confidence, market valuation and intellectual property when large amounts of data are stolen. In this research project we will use an enhanced “screen and review” approach to combating exfiltration in a large data set of activity logs within a large corporate network.
The Métis Nation of BC, (MNBC) has been challenged with a sense of division among staff, governance and citizenship. As a result, programming is missing a strategic framework while ministries and chartered communities operate as independent silos. There is a clear and collective desire to rejuvenate Métis culture but without first acknowledging current state and a clear future state (Stroh, 2015), it is difficult to coordinate initiatives within all the cohesive Métis groups in BC.
In AI safety, compliance ensures that a model adheres to operational specifications at runtime to avoid adverse events for the end user. This proposal looks at obtaining model compliance in two ways: (i) applying corrective measures to a non-compliant Machine Learning (ML) model and (ii) ensuring compliance throughout the model’s training process. We aim to achieve the first via removal of gradient information related to features involved in biasing the model.
The Fundamental Review of the Trading Book (FRTB) is a set of regulations by the Basel committee, which is expected to be implemented by banks by 2022. The regulation targets market risk management in banking industry. According to FRTB, banks need to post extra capital against non-modellable risk factors, which could account for 30% of total market risk capital requirement. Reducing the weight of non-modellable risk factors can greatly reduce the required capital and thus increase banks’ profitability.
This project is about implementing a technique called Agent-Based modelling (ABM) so it can work better in real-world application. Particularly, it aims to help policy makers to do more adaptive decisions when the whole economics environment changes. For example, how to set the federal interest rate after COVID-19 panic? This model could simulate how all kinds of people, regulators, corporations, banks, or investors interact with others and how that interaction could cause specific things to happen to them and to the market more broadly.
A geospatial query is a question where the concept of location is necessary for formulating the answer. Furthermore, we are not simply interested in spatial relationships, but also with the ways in which people can possibly move through space given the goals that they want to achieve. We therefore want to predict the behaviour of people moving through urban environments based on observations about their purchases. In this project, we will explore how can models of commonsense knowledge can be used for automated reasoning to answer geospatial queries and to infer consumer behaviour.
