ML Anthology

Bellman Goes Relational

Kristian Kersting, Martijn van Otterlo, Luc De Raedt
ICML 2004
doi:10.1145/1015330.1015401 /icml/2004/kersting2004icml-bellman/

Abstract

Motivated by the interest in relational reinforcement learning, we introduce a novel relational Bellman update operator called ReBel. It employs a constraint logic programming language to compactly represent Markov decision processes over relational domains. Using ReBel, a novel value iteration algorithm is developed in which abstraction (over states and actions) plays a major role. This frameworkprovides new insights into relational reinforcement learning. Convergence results as well as experiments are presented.

PDF ICML Semantic Scholar

Cite

Text

Kersting et al. "Bellman Goes Relational." International Conference on Machine Learning, 2004. doi:10.1145/1015330.1015401

Markdown

[Kersting et al. "Bellman Goes Relational." International Conference on Machine Learning, 2004.](https://mlanthology.org/icml/2004/kersting2004icml-bellman/) doi:10.1145/1015330.1015401

BibTeX

@inproceedings{kersting2004icml-bellman,
  title     = {{Bellman Goes Relational}},
  author    = {Kersting, Kristian and van Otterlo, Martijn and De Raedt, Luc},
  booktitle = {International Conference on Machine Learning},
  year      = {2004},
  doi       = {10.1145/1015330.1015401},
  url       = {https://mlanthology.org/icml/2004/kersting2004icml-bellman/}
}