PolicyBlocks: An Algorithm for Creating Useful Macro-Actions in Reinforcement Learning

Abstract

We present PolicyBlocks, an algorithm by which a reinforcement learning agent can extract useful macro-actions from a set of related tasks. The agent creates macroactions by finding commonalities in solutions to previous tasks. Using these macro-actions, learning to do future related tasks is accelerated.

Cite

Text

Pickett and Barto. "PolicyBlocks: An Algorithm for Creating Useful Macro-Actions in Reinforcement Learning." International Conference on Machine Learning, 2002.

Markdown

[Pickett and Barto. "PolicyBlocks: An Algorithm for Creating Useful Macro-Actions in Reinforcement Learning." International Conference on Machine Learning, 2002.](https://mlanthology.org/icml/2002/pickett2002icml-policyblocks/)

BibTeX

@inproceedings{pickett2002icml-policyblocks,
  title     = {{PolicyBlocks: An Algorithm for Creating Useful Macro-Actions in Reinforcement Learning}},
  author    = {Pickett, Marc and Barto, Andrew G.},
  booktitle = {International Conference on Machine Learning},
  year      = {2002},
  pages     = {506-513},
  url       = {https://mlanthology.org/icml/2002/pickett2002icml-policyblocks/}
}