ML Anthology

Incremental Version-Space Merging

Haym Hirsh
ICML 1990 pp. 330-338
doi:10.1016/B978-1-55860-141-3.50043-2 /icml/1990/hirsh1990icml-incremental/

Abstract

This paper describes a generalization of Mitchell's version-space approach to concept learning that significantly extends its range of applicability. The key idea is to remove the central idea of a version space being the set of concepts strictly consistent with training data, and allow arbitrary sets of concepts, however generated, as long as they can be represented by boundary sets. Learning is accomplished with version space intersection, rather than the traditional candidate-elimination algorithm. Applications of the learning method, incremental version-space merging, include learning from forms of inconsistent data and combining empirical and analytical learning.

ICML Semantic Scholar

Cite

Text

Hirsh. "Incremental Version-Space Merging." International Conference on Machine Learning, 1990. doi:10.1016/B978-1-55860-141-3.50043-2

Markdown

[Hirsh. "Incremental Version-Space Merging." International Conference on Machine Learning, 1990.](https://mlanthology.org/icml/1990/hirsh1990icml-incremental/) doi:10.1016/B978-1-55860-141-3.50043-2

BibTeX

@inproceedings{hirsh1990icml-incremental,
  title     = {{Incremental Version-Space Merging}},
  author    = {Hirsh, Haym},
  booktitle = {International Conference on Machine Learning},
  year      = {1990},
  pages     = {330-338},
  doi       = {10.1016/B978-1-55860-141-3.50043-2},
  url       = {https://mlanthology.org/icml/1990/hirsh1990icml-incremental/}
}