About
Appointed Canada CIFAR AI Chair – 2021
Marlos C. Machado is an assistant professor at the University of Alberta, a Canada CIFAR AI Chair and Fellow at Amii, and a principal investigator in the Reinforcement Learning and Artificial Intelligence (RLAI) group.
Machado’s research focuses on sequential decision-making problems and he is particularly interested in developing approaches that learn continually and that are able to create abstractions of their observations and of their behaviours. For continual learning, he has been revisiting some of the fundamentals of deep reinforcement learning to allow its algorithms to actually learn continually. For abstractions of observations, he has shown the benefits of incorporating the sequential structure of decision making problems into the representation learning process, as well as the benefits of augmenting agents’ representations with predictions and uncertainty estimates. For temporal abstractions, Marlos has introduced the idea of using learned representations to discover options (i.e., courses of actions) for temporally extended exploration, an approach effective in both episodic and continual settings. Several of these ideas were incorporated in the design of a deep RL algorithm for controlling balloons in the stratosphere; one of the first deployments of deep RL in the real-world. His research has been featured in popular media such as BBC, Bloomberg TV, The Verge, and Wired.
Awards
- Best Paper Honorable Mention, AISTATS, 2023
- Best Paper Award, ICML Workshop on Exploration in RL, 2019
- Best Paper Award, ICML Workshop on Exploration in RL, 2018
- Best Paper Honorable Mention, AAMAS, 2016
Relevant Publications
- Gomez, D., Bowling, M., Machado, M.C. (2024). Proper Laplacian Representation Learning. International Conference on Representation Learning.
- Machado, M.C., Barreto, A., Precup, D., & Bowling, M. (2023). Temporal Abstraction in Reinforcement Learning with the Successor Representation. Journal of Machine Learning Research (JMLR), 24(80):1--69.
- Abbas, Z., Zhao, R., Modayil, J., White, A., & Machado, M.C. (2023). Loss of Plasticity in Continual Deep Reinforcement Learning. Conference on Lifelong Learning Agents, 2023
Bellemare, M., Candido, S., Castro, P., Gong, J., Machado, M. C., Moitra, S., Ponda, S., & Wang, Z. (2020). Autonomous Navigation of Stratospheric Balloons using Reinforcement Learning. Nature; 588:77‑82.
Machado, M. C., Bellemare, M., Talvitie, E., Veness, J., Hausknecht, M., Bowling, M. (2018). Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents. Journal of Artificial Intelligence Research 61: 523‑562.