English

Learning States Representations in POMDP

Machine Learning 2014-06-18 v4

Abstract

We propose to deal with sequential processes where only partial observations are available by learning a latent representation space on which policies may be accurately learned.

Cite

@article{arxiv.1312.6042,
  title  = {Learning States Representations in POMDP},
  author = {Gabriella Contardo and Ludovic Denoyer and Thierry Artieres and Patrick Gallinari},
  journal= {arXiv preprint arXiv:1312.6042},
  year   = {2014}
}

Comments

4 pages

R2 v1 2026-06-22T02:32:48.562Z