Multiple Graduate Student Positions in Reinforcement Learning

[Amir-massoud Farahmand]

Dear all,

Several faculty members of the Vector Institute in Canada work on reinforcement learning, sequential decision making, and closely related research areas. Each of us is affiliated with one of the Canadian universities, but we are all affiliated with the Vector Institute and often collaborate with each other. Most of our students are located at the Vector Institute, which is a thriving environment for machine learning research.

We plan to recruit several graduate students this year. If you are interested in reinforcement learning research, please apply through our respective departments. We welcome both domestic and international students.

The name of faculty members who work on RL-related topics are as follows. Please check their webpages to find the best match based on your interests, and particular instructions that they may have for prospective students.

• Amir-massoud Farahmand (Department of Computer Science, University of Toronto). Interests: Theoretical RL, Model-based RL, Risk and Robustness (Prospective Students)

• Angela Schoellig (Institute for Aerospace Studies, University of Toronto). Interests: Robot Control and Learning; Reinforcement Learning for Robotics; Mobile Manipulation; Self Driving and Flying

• Animesh Garg (Department of Computer Science, University of Toronto). Interests: Generalizable Autonomy for Robotics, Reinforcement Learning, Optimal Control, Causal Decision Making, Neural Architectures for Decision Making (Prospective Students)

• Daniel M. Roy (Department of Statistical Sciences and Computer Science, University of Toronto). Interests: Theory for worst case and adaptive online learning, bandits, and, in the future, RL. Prospective students should apply to both CS and Statistics, if appropriate.
  

• Florian Shkurti (Department of Computer Science, University of Toronto). Interests: Machine learning for planning and control, Robotics, Inverse RL, Imitation Learning

• Jeff Clune (Computer Science, University of British Columbia). Interests: Deep Reinforcement Learning, AI-Generating Algorithms

• Joseph J. Williams (Departments of Computer Science, Statistical Sciences, and Psychology, University of Toronto). Interests: Multi-armed bandits for healthcare and education (Prospective Students)

• Pascal Poupart (School of Computer Science, University of Waterloo). Interests: Partially Observable Reinforcement Learning, Bayesian Reinforcement Learning, Causal Reinforcement Learning, Federated Reinforcement Learning, Object-Oriented Reinforcement Learning, Reinforcement Learning in Natural Language Processing

• Scott Sanner (Department of Mechanical and Industrial Engineering, Cross-appointed in Department of Computer Science, University of Toronto). Interests: Data-driven Decision Making, Sequential Decision Optimization

• Sheila McIlraith (Department of Computer Science, University of Toronto). Interests: Sequential decision making and reinforcement learning, Program synthesis, Human-compatible AI

Please note that, because of the high volume of inquiries, some of the listed faculty may not be able to respond to individual emails from prospective students. This should not be interpreted as a lack of interest. It is sufficient to mention their names in your application, and they will closely look at your application.

Each department has its own webpage, admission deadline and requirements. Please check them for the updated information.

• University of Toronto

  • Computer Science (Admission – Deadline: December 1)

  • Mechanical and Industrial Engineering (Admission – Deadline: Jan 1, 2022)

  • University of Toronto Institute for Aerospace Studies (Admission – Deadline: December 15, 2021 (fee); January 15, 2022 (application material))

• University of Waterloo (Computer Science) (Admission – Deadline: December 15)

• University of British Columbia (Computer Science) (Admission – Deadline: December 15) 

In addition to these faculty members whose work has a significant focus on RL, there are many other researchers at the Vector Institute who work on other aspects of machine learning and deep learning, including fundamental algorithm design, representation learning, generative models, optimization, theory of ML and DL, computer vision, privacy, fairness, intersection of quantum computing and ML, and applications in healthcare, material design, music, etc. You can find their names here.

About Vector Institute

The Vector Institute is an independent non-profit corporation, with Faculty Members and Affiliates from the University of Toronto, University of Waterloo, University of Guelph, Dalhousie University, and other Canadian universities. It is supported with generous funding from the provincial and federal governments, as well as Canadian industry sponsors. The Vector Institute is located in the MaRS Discovery District building, spanning nearly the entire 7th floor and overlooking downtown Toronto and the beautiful Queen’s Park. On any given day, the Vector Institute houses over a hundred students, dozens of Faculty Members, supported with state-of-the-art compute power, and dedicated professional staff. The daily life and concentration of expertise in the Institute fosters collaboration and the exchange of ideas among its members through talks, seminar series, visitors, and tutorials. The Vector Institute’s vision is to drive excellence in the creation of artificial intelligence, to use it to foster economic growth, and to improve the lives of Canadians. To that end, the Vector Institute has close ties to both academia and industry.

Best Regards,

Amir-massoud Farahmand on behalf of my colleagues

[Amir-massoud Farahmand]

I missed adding one of our faculty members who work on RL:

• Jimmy Ba (Department of Computer Science, University of Toronto). Interests: Neural Networks, Deep Learning, Reinforcement Learning