ML Anthology

Using Inverse Resolution to Learn Relations from Experiments

David Humme, Claude Sammut
ICML 1991 pp. 412-416
doi:10.1016/B978-1-55860-200-7.50085-4 /icml/1991/humme1991icml-using/

Abstract

We are concerned with learning relations in a reactive environment. A learning agent observes sequences of actions that may change the relationships of objects in the world. The observed sequence is used to form a theory which can be generalised and tested by experimentation. The section of an experiment is dependent upon the current state of the world. Therefore, the learning algorithm must be “opportunistic― in the sense that it attempts to test hypotheses as circumstances permit.

PDF ICML Semantic Scholar

Cite

Text

Humme and Sammut. "Using Inverse Resolution to Learn Relations from Experiments." International Conference on Machine Learning, 1991. doi:10.1016/B978-1-55860-200-7.50085-4

Markdown

[Humme and Sammut. "Using Inverse Resolution to Learn Relations from Experiments." International Conference on Machine Learning, 1991.](https://mlanthology.org/icml/1991/humme1991icml-using/) doi:10.1016/B978-1-55860-200-7.50085-4

BibTeX

@inproceedings{humme1991icml-using,
  title     = {{Using Inverse Resolution to Learn Relations from Experiments}},
  author    = {Humme, David and Sammut, Claude},
  booktitle = {International Conference on Machine Learning},
  year      = {1991},
  pages     = {412-416},
  doi       = {10.1016/B978-1-55860-200-7.50085-4},
  url       = {https://mlanthology.org/icml/1991/humme1991icml-using/}
}