Publications

Sort by: [year] [type] [author]

Manifold Representations for Value-Function Approximation

Robert Glaubius and William D. Smart.
In "Learning and Planning in Markov Processes - Advances and Challenges: Papers from the 2004 AAAI Workshop", Daniela Pucci de Farias, Shie Mannor, Doina Precup, and Georgios Theocharous (eds)., pages 13-18, June 2004.

Available in AAAI Technical Report WS-04-08.

We present a novel approach to modeling the reinforcement learning value function using a manifold representation. By explicitly modeling the topology of the value function domain, traditional problems with discontinuities and resolution can be addressed without resorting to complex function approximators. We describe the mathematical underpinnings of our approach, and show how manifold techniques can be applied to value-function approximation. We also present techniques for constructing a manifold representation of the domain, and show their effectiveness on example problems.

Paper: [PDF]
External link: [link]

@inproceedings{aaai2004a,
  author = {Glaubius, Robert and Smart, William D.},
  editor = {de Farias, Daniela Pucci and Mannor, Shie and Precup, Doina and Theocharous, Georgios},
  title = {Manifold Representations for Value-Function Approximation},
  booktitle = {Learning and Planning in Markov Processes --- Advances and Challenges: Papers from the 2004 {AAAI} Workshop},
  pages = {13--18},
  note = {Available in {AAAI} Technical Report {WS-04-08}.},
  month = {June},
  year = {2004}
}