Alberta Machine Intelligence Institute

Fellow & Canada CIFAR AI Chair

Marlos C. Machado

Academic Affiliations

Assistant Professor – University of Alberta (Computing Science)

Focus

Deep reinforcement learning; representation learning; continual learning

Marlos’ research focuses on machine learning, with a particular emphasis on deep reinforcement learning, representation learning, and continual learning, alongside their real-world applications.

Representation and reinforcement learning

Through his research, Marlos seeks to develop reinforcement learning methods that can be meaningfully used in real-world settings. His research includes several different avenues, including designing representation learning methods tailored for reinforcement learning problems, developing AI agents that are capable of discovering temporally extended-behaviors (known as options), and creating systems capable of continual learning. Marlos is also passionate about reproducibility and proper experimentation in machine learning, having led several efforts on this topic in the past.

Marlos completed his B.Sc. and M.Sc. at UFMG, Brazil, and his Ph.D. at the University of Alberta. During my Ph.D., among other things, he introduced stochasticity and game modes in the popular Arcade Learning Environment, and popularized the idea of temporally-extended exploration through options, introducing the idea of eigenoptions. Before becoming a professor, he was a researcher at DeepMind and at Google Brain for four years; during which time he made several contributions to reinforcement learning, including the application of deep reinforcement learning to control Loon’s stratospheric balloons.

His work has been published in the leading conferences and journals in machine learning, including Nature, JMLR, JAIR, NeurIPS, ICML and ICLR. His research has also been featured in popular media such as the BBC, Bloomberg TV, The Verge, and Wired.

Latest News