ML Anthology

Revision Cost for Theory Refinement

Rei Hamakawa
ICML 1991 pp. 514-518
doi:10.1016/B978-1-55860-200-7.50105-7 /icml/1991/hamakawa1991icml-revision/

Abstract

This paper describes a new method for incremental refinement of the approximate domain theory. The method presented here is based a concept of on revision cost which is a measure of tree transformation. The method is applied to a new system of integrated learning, which combines empirical learning and explanation-based learning. This approach allows handling refinement procedures in a uniform way and changes in the refinement strategy can be made by changing cost parameters according to the performance achieved.

ICML Semantic Scholar

Cite

Text

Hamakawa. "Revision Cost for Theory Refinement." International Conference on Machine Learning, 1991. doi:10.1016/B978-1-55860-200-7.50105-7

Markdown

[Hamakawa. "Revision Cost for Theory Refinement." International Conference on Machine Learning, 1991.](https://mlanthology.org/icml/1991/hamakawa1991icml-revision/) doi:10.1016/B978-1-55860-200-7.50105-7

BibTeX

@inproceedings{hamakawa1991icml-revision,
  title     = {{Revision Cost for Theory Refinement}},
  author    = {Hamakawa, Rei},
  booktitle = {International Conference on Machine Learning},
  year      = {1991},
  pages     = {514-518},
  doi       = {10.1016/B978-1-55860-200-7.50105-7},
  url       = {https://mlanthology.org/icml/1991/hamakawa1991icml-revision/}
}