Three faculty members from the University of Toronto’s Department of Computer Science are recipients of Early Researcher Awards from the Government of Ontario in support of various artificial intelligence projects.
Assistant Professor Amir-massoud Farahmand, Associate Professor Gennady Pekhimenko and Assistant Professor Bo Wang have each been awarded funding in the latest round of the competitive program announced in March.
Early Researcher Awards are aimed at “driving innovation across the province, helping researchers to build research teams and improving Ontario’s ability to attract and retain the best and brightest talent in a variety of disciplines.”
Here is a closer look at the awarded projects:
Amir-massoud Farahmand
Assistant Professor, Department of Computer Science
Project Title: Accelerated Reinforcement Learning Algorithms
Reinforcement learning (RL) algorithms train agents such as robots, autonomous cars or health-care assistants to make decisions with long-term consequences and plan far into the future. However, this long-term planning requires a lot of computing power and data. This research aims to rethink computational mechanisms in RL and develop faster methods for making long-term planning more computational and data efficient, paving the way for this subfield of machine learning to be more applicable to real-world problems.
Gennady Pekhimenko
Associate Professor, UTSC Department of Computer and Mathematical Sciences and tri-campus graduate Department of Computer Science
Project Title: Computer Systems for Efficient and Scalable Machine Learning
New machine learning (ML) models and algorithms are developing at a rapid pace, but they need well-designed computer systems to reach their full potential. This project sets out to bridge a gap in Ontario’s ML landscape with focused research on computer systems and architecture. By integrating insights on algorithms, systems and hardware, this project aims to make ML training more energy efficient and scalable.
Bo Wang
Assistant Professor, Department of Laboratory Medicine and Pathobiology and Department of Computer Science
Project Title: Prediction of heart failure with interpretable deep learning
Heart failure is the most common cause of hospitalization for Ontarians over the age of 65 and is ultimately fatal for 50 per cent of patients within five years of diagnosis. Early diagnosis and intervention are crucial to improve survivability and quality of life, but current tools are insufficient for a timely response. With the influx of routinely collected administrative and clinical health-care data, this project will leverage AI tools to develop an early detection system for patients at high risk of heart failure.