ML Anthology

Concept Sharing: A Means to Improve Multi-Concept Learning

Piew Datta, Dennis F. Kibler
ICML 1993 pp. 89-96
doi:10.1016/B978-1-55860-307-3.50018-6 /icml/1993/datta1993icml-concept/

Abstract

This paper describes several means for sharing between related concepts to improve learning in the same domain. The sharing comes in the form of substructures or possibly entire structures of previous concepts which may aid in learning other concepts. These substructures highlight useful information in the domain. Using two domains, we evaluate the effectiveness of concept sharing with respect to accuracy, concept size, search complexity, and noise resistance.

ICML Semantic Scholar

Cite

Text

Datta and Kibler. "Concept Sharing: A Means to Improve Multi-Concept Learning." International Conference on Machine Learning, 1993. doi:10.1016/B978-1-55860-307-3.50018-6

Markdown

[Datta and Kibler. "Concept Sharing: A Means to Improve Multi-Concept Learning." International Conference on Machine Learning, 1993.](https://mlanthology.org/icml/1993/datta1993icml-concept/) doi:10.1016/B978-1-55860-307-3.50018-6

BibTeX

@inproceedings{datta1993icml-concept,
  title     = {{Concept Sharing: A Means to Improve Multi-Concept Learning}},
  author    = {Datta, Piew and Kibler, Dennis F.},
  booktitle = {International Conference on Machine Learning},
  year      = {1993},
  pages     = {89-96},
  doi       = {10.1016/B978-1-55860-307-3.50018-6},
  url       = {https://mlanthology.org/icml/1993/datta1993icml-concept/}
}