A
Andre Barreto
Researcher at Google
Publications - 86
Citations - 2306
Andre Barreto is an academic researcher from Google. The author has contributed to research in topics: Reinforcement learning & Computer science. The author has an hindex of 21, co-authored 80 publications receiving 1695 citations. Previous affiliations of Andre Barreto include Federal University of Rio de Janeiro & McGill University.
Papers
More filters
Proceedings Article
Successor Features for Transfer in Reinforcement Learning
Andre Barreto,Will Dabney,Rémi Munos,Jonathan J. Hunt,Tom Schaul,Hado van Hasselt,David Silver +6 more
TL;DR: This work proposes a transfer framework for the scenario where the reward function changes between tasks but the environment's dynamics remain the same, and derives two theorems that set the approach in firm theoretical ground and presents experiments that show that it successfully promotes transfer in practice.
Posted Content
Successor Features for Transfer in Reinforcement Learning
Andre Barreto,Will Dabney,Rémi Munos,Jonathan J. Hunt,Tom Schaul,Hado van Hasselt,David Silver +6 more
TL;DR: This article proposed a transfer framework for the scenario where the reward function changes between tasks but the environment's dynamics remain the same, and derived two theorems that set their approach in firm theoretical ground and present experiments that show that it successfully promotes transfer in practice, significantly outperforming alternative methods in a sequence of navigation tasks and in the control of a simulated robotic arm.
Posted Content
The Predictron: End-To-End Learning and Planning
David Silver,Hado van Hasselt,Matteo Hessel,Tom Schaul,Arthur Guez,Tim Harley,Gabriel Dulac-Arnold,David P. Reichert,Neil C. Rabinowitz,Andre Barreto,Thomas Degris +10 more
TL;DR: In this paper, the authors introduce the predictron architecture, which consists of a fully abstract model, represented by a Markov reward process, that can be rolled forward multiple "imagined" planning steps.
Proceedings Article
The predictron: end-to-end learning and planning
David Silver,Hado van Hasselt,Matteo Hessel,Tom Schaul,Arthur Guez,Tim Harley,Gabriel Dulac-Arnold,David P. Reichert,Neil C. Rabinowitz,Andre Barreto,Thomas Degris +10 more
TL;DR: The predictron consists of a fully abstract model, represented by a Markov reward process, that can be rolled forward multiple "imagined" planning steps that accumulates internal rewards and values over multiple planning depths.
Proceedings Article
Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement
Andre Barreto,Diana Borsa,John Quan,Tom Schaul,David Silver,Matteo Hessel,Daniel J. Mankowitz,Augustin Zídek,Rémi Munos +8 more
TL;DR: This paper shows that the transfer promoted by SFs & GPI leads to very good policies on unseen tasks almost instantaneously, and describes how to learn policies specialised to the new tasks in a way that allows them to be added to the agent's set of skills, and thus be reused in the future.