A
Anders Jonsson
Researcher at Pompeu Fabra University
Publications - 113
Citations - 1691
Anders Jonsson is an academic researcher from Pompeu Fabra University. The author has contributed to research in topics: Reinforcement learning & Markov decision process. The author has an hindex of 19, co-authored 105 publications receiving 1259 citations. Previous affiliations of Anders Jonsson include Information Technology University & University of Massachusetts Amherst.
Papers
More filters
Posted Content
A unified view of entropy-regularized Markov decision processes
TL;DR: A general framework for entropy-regularized average-reward reinforcement learning in Markov decision processes (MDPs) is proposed, showing that using the conditional entropy of the joint state-action distributions as regularization yields a dual optimization problem closely resembling the Bellman optimality equations.
Journal Article
Causal Graph Based Decomposition of Factored MDPs
Anders Jonsson,Andrew G. Barto +1 more
TL;DR: Experimental results show that the decomposition introduced by VISA can significantly accelerate construction of an optimal, or near-optimal, policy.
Proceedings Article
Automated State Abstraction for Options using the U-Tree Algorithm
Anders Jonsson,Andrew G. Barto +1 more
TL;DR: This paper adapting McCallum's U-Tree algorithm to automatically build option-specific representations of the state feature space, and illustrating the resulting algorithm using a simple hierarchical task suggests that automated option- specific state abstraction is an attractive approach to making hierarchical learning systems more effective.
Posted Content
Fast active learning for pure exploration in reinforcement learning
Pierre Ménard,Omar Darwiche Domingues,Anders Jonsson,Emilie Kaufmann,Edouard Leurent,Michal Valko +5 more
TL;DR: It is shown that, surprisingly, for a pure-exploration objective of reward-free exploration, bonuses that scale with 1/n bring faster learning rates, improving the known upper bounds with respect to the dependence on the horizon H.
Journal ArticleDOI
The complexity of planning problems with simple causal graphs
Omer Giménez,Anders Jonsson +1 more
TL;DR: A polynomial-time algorithm that uses macros to generate plans for the class 3S of planning problems with binary state variables and acyclic causal graphs implies that plan generation may be tractable even when a planning problem has an exponentially long minimal solution.