Exemplar-Based Theory Rejection: An Approach to the Experience Consistency Problem
Abstract
Theory revision is the process of extending a domain theory to accommodate anomalies. Of primary concern in theory revision is the experience consistency problem–the problem of insuring that the candidate revised theories are consistent with the previous experience of the system. This paper describes an approach called exemplar-based theory rejection a solution to the experience consistency problem. Exemplar-based theory rejection collects and maintains a representative set of examples, called exemplars, of the use of the components of the theory in successful reasoning tasks. These exemplars are used to test the candidate revised theories generated by theory revision. Exemplar-based theory rejection has been implemented in the COAST system–a system that revises qualitative theories of the physical world.
Cite
Text
Rajamoney. "Exemplar-Based Theory Rejection: An Approach to the Experience Consistency Problem." International Conference on Machine Learning, 1989. doi:10.1016/B978-1-55860-036-2.50078-3Markdown
[Rajamoney. "Exemplar-Based Theory Rejection: An Approach to the Experience Consistency Problem." International Conference on Machine Learning, 1989.](https://mlanthology.org/icml/1989/rajamoney1989icml-exemplar/) doi:10.1016/B978-1-55860-036-2.50078-3BibTeX
@inproceedings{rajamoney1989icml-exemplar,
title = {{Exemplar-Based Theory Rejection: An Approach to the Experience Consistency Problem}},
author = {Rajamoney, Shankar A.},
booktitle = {International Conference on Machine Learning},
year = {1989},
pages = {284-289},
doi = {10.1016/B978-1-55860-036-2.50078-3},
url = {https://mlanthology.org/icml/1989/rajamoney1989icml-exemplar/}
}