Corpus ID: 7183484

Hierarchical Reinforcement Learning Based on Subgoal Discovery and Subpolicy Specialization

@inproceedings{Bakker2005HierarchicalRL,
  title={Hierarchical Reinforcement Learning Based on Subgoal Discovery and Subpolicy Specialization},
  author={B. Bakker},
  year={2005}
}
A method is described for hierarchical reinforcement learning. High-level policies automatically discover subgoals; low-level policies learn to specialize for different subgoals. Subgoals are represented as desired abstract observations which cluster raw input data. Experiments shows that this method outperforms several flat methods. The full paper can be found in [1]. Introduction. The promise of scaling up reinforcement learning (RL) through hierarchical RL (HRL) is widely acknowledged. The… Expand
132 Citations
Learning Representations in Model-Free Hierarchical Reinforcement Learning
  • 12
  • PDF
Grounding subgoals in information transitions
  • S. V. Dijk, D. Polani
  • Computer Science
  • 2011 IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning (ADPRL)
  • 2011
  • 28
  • PDF
EARNING WITH P OLICY S KETCHES
Learning to generalize and reuse skills using approximate partial policy homomorphisms
  • S. Rajendran, M. Huber
  • Computer Science
  • 2009 IEEE International Conference on Systems, Man and Cybernetics
  • 2009
  • 8
  • PDF
Generalizing and Categorizing Skills in Reinforcement Learning Agents Using Partial Policy Homomorphisms
  • 2
Hierarchical Approaches
  • B. Hengst
  • Computer Science
  • Reinforcement Learning
  • 2012
  • 18
  • Highly Influenced
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 15 REFERENCES
Hierarchical Reinforcement Learning with the MAXQ Value Function Decomposition
  • 1,327
  • PDF
Between MDPs and Semi-MDPs: A Framework for Temporal Abstraction in Reinforcement Learning
  • 2,386
  • PDF
Automatic Discovery of Subgoals in Reinforcement Learning using Diverse Density
  • 437
  • PDF
Recent Advances in Hierarchical Reinforcement Learning
  • 425
Feudal Reinforcement Learning
  • 541
  • Highly Influential
  • PDF
Self-segmentation of sequences: automatic formation of hierarchies of sequential behaviors
  • 50
Reinforcement Learning: A Survey
  • 6,701
  • Highly Influential
  • PDF
Multi-Agent Residual Advantage Learning with General Function Approximation.
  • 40
HQ-Learning
  • 177
Multi-player residual advantage learning with general function
  • 40
  • PDF
...
1
2
...