Marlos C. Machado
About
Marlos C. Machado is an Assistant Professor in the Department of Computing Science at the University of Alberta, a Canada CIFAR AI Chair and Amii Fellow, and a Principal Investigator in the Reinforcement Learning & Artificial Intelligence (RLAI) group. His research focuses on deep reinforcement learning, representation learning, continual learning, and real-world applications—emphasizing algorithms that autonomously discover temporal abstractions (“eigenoptions”) via the successor representation. During his Ph.D., he introduced stochasticity and game modes in the Arcade Learning Environment and popularized eigenoptions; post-Ph.D., he spent four years at DeepMind and Google Brain, contributing to real-world systems like stratospheric balloon control. His work, featured in top venues such as Nature, JMLR, NeurIPS, ICML, and ICLR, as well as media outlets like BBC, Bloomberg, The Verge, and Wired, continues to influence foundational advances in exploration and abstraction in RL. Read more