@article{directedpolicygradientforsafe,
  title = {Directed Policy Gradient for Safe Reinforcement Learning with Human Advice},
  author = {Hélène Plisnier and Denis Steckelmacher and Tim Brys and Diederik M. Roijers and Ann Nowé},
  year = {2018},
  eprint = {1808.04096},
  archivePrefix = {arXiv},
  url = {http://arxiv.org/abs/1808.04096v1},
}