English

Variance Reduction Methods for Sublinear Reinforcement Learning

Artificial Intelligence 2020-06-30 v5 Machine Learning Machine Learning

Abstract

There is a technical issue in the analysis that is not easily fixable. We, therefore, withdraw the submission. Sorry for the inconvenience.

Cite

@article{arxiv.1802.09184,
  title  = {Variance Reduction Methods for Sublinear Reinforcement Learning},
  author = {Sham Kakade and Mengdi Wang and Lin F. Yang},
  journal= {arXiv preprint arXiv:1802.09184},
  year   = {2020}
}

Comments

There is a technical issue in the analysis that is not easily fixable