Structure in the Space of Value Functions

@article{Foster2002StructureIT,
  title={Structure in the Space of Value Functions},
  author={David J. Foster and Peter Dayan},
  journal={Machine Learning},
  year={2002},
  volume={49},
  pages={325-346}
}
Solving in an efficient manner many different optimal control tasks within the same underlying environment requires decomposing the environment into its computationally elemental fragments. We suggest how to find fragmentations using unsupervised, mixture model, learning methods on data derived from optimal value functions for multiple tasks, and show that these fragmentations are in accord with observable structure in the environments. Further, we present evidence that such fragments can be of… CONTINUE READING
Highly Cited
This paper has 87 citations. REVIEW CITATIONS
33 Citations
38 References
Similar Papers

Citations

Publications citing this paper.

87 Citations

051015'05'08'11'14'17
Citations per Year
Semantic Scholar estimates that this publication has 87 citations based on the available data.

See our FAQ for additional information.

Similar Papers

Loading similar papers…