ML Anthology

Fuzzy Substructure Discovery

Lawrence B. Holder, Diane J. Cook, Horst Bunke
ICML 1992 pp. 218-223
doi:10.1016/B978-1-55860-247-2.50033-4 /icml/1992/holder1992icml-fuzzy/

Abstract

This paper describes a method for discovering substructures in data using a fuzzy graph match. A previous implementation of the Subdue system discovers substructures based on the psychologically-motivated criteria of cognitive savings, compactness, connectivity and coverage. However, the instances in the data must exactly match the discovered substructures. We describe a new implementation of Subdue that employs a fuzzy graph match to discover substructures which occur often in the data, but not always in the same form. This fuzzy substructure discovery can be used to formulate fuzzy concepts, compress the data description, and discover interesting structures in data that are found either in their pure form or in a slightly convoluted form. Examples from the domains of scene analysis and chemical compound analysis demonstrate the fuzzy discovery technique.

ICML Semantic Scholar

Cite

Text

Holder et al. "Fuzzy Substructure Discovery." International Conference on Machine Learning, 1992. doi:10.1016/B978-1-55860-247-2.50033-4

Markdown

[Holder et al. "Fuzzy Substructure Discovery." International Conference on Machine Learning, 1992.](https://mlanthology.org/icml/1992/holder1992icml-fuzzy/) doi:10.1016/B978-1-55860-247-2.50033-4

BibTeX

@inproceedings{holder1992icml-fuzzy,
  title     = {{Fuzzy Substructure Discovery}},
  author    = {Holder, Lawrence B. and Cook, Diane J. and Bunke, Horst},
  booktitle = {International Conference on Machine Learning},
  year      = {1992},
  pages     = {218-223},
  doi       = {10.1016/B978-1-55860-247-2.50033-4},
  url       = {https://mlanthology.org/icml/1992/holder1992icml-fuzzy/}
}