ML Anthology

The Flexibility of Speculative Refinement

Susan Craw, Derek H. Sleeman
ICML 1991 pp. 28-32
doi:10.1016/B978-1-55860-200-7.50010-6 /icml/1991/craw1991icml-flexibility/

Abstract

KRUST is an automated refiner which speculatively generates many possible refinements for a knowledge base (KB) to correct a single wrongly solved task. It selects a single refined KB which performs best empirically. This philosophy allows a flexible approach to refinement, because we have an opportunity to tailor the filtering and empirical testing towards particular aims. The focus of this paper is the reconstruction of KBs.

ICML Semantic Scholar

Cite

Text

Craw and Sleeman. "The Flexibility of Speculative Refinement." International Conference on Machine Learning, 1991. doi:10.1016/B978-1-55860-200-7.50010-6

Markdown

[Craw and Sleeman. "The Flexibility of Speculative Refinement." International Conference on Machine Learning, 1991.](https://mlanthology.org/icml/1991/craw1991icml-flexibility/) doi:10.1016/B978-1-55860-200-7.50010-6

BibTeX

@inproceedings{craw1991icml-flexibility,
  title     = {{The Flexibility of Speculative Refinement}},
  author    = {Craw, Susan and Sleeman, Derek H.},
  booktitle = {International Conference on Machine Learning},
  year      = {1991},
  pages     = {28-32},
  doi       = {10.1016/B978-1-55860-200-7.50010-6},
  url       = {https://mlanthology.org/icml/1991/craw1991icml-flexibility/}
}