ML Anthology

Issues in the Design of Operator Composition Systems

Steven Minton
ICML 1990 pp. 304-312
doi:10.1016/B978-1-55860-141-3.50040-7 /icml/1990/minton1990icml-issues/

Abstract

Many learning problem solvers operate by composing operator sequences, so that a learned sequence can be applied as a unit during subsequent problem solving. In this paper we will describe an abstract model of the operator composition and problem solving processes, and use the model to analyze severed design issues that affect the utility of the learning method. We will focus primarily on design issues that arose during the implementation of the PRODIGY system[16; 18] and two of its predecessors[17; 15]. The purpose of this paper is to consider these issues from the common perspective offered by our model, and to summarize the relevant research in the field.

ICML Semantic Scholar

Cite

Text

Minton. "Issues in the Design of Operator Composition Systems." International Conference on Machine Learning, 1990. doi:10.1016/B978-1-55860-141-3.50040-7

Markdown

[Minton. "Issues in the Design of Operator Composition Systems." International Conference on Machine Learning, 1990.](https://mlanthology.org/icml/1990/minton1990icml-issues/) doi:10.1016/B978-1-55860-141-3.50040-7

BibTeX

@inproceedings{minton1990icml-issues,
  title     = {{Issues in the Design of Operator Composition Systems}},
  author    = {Minton, Steven},
  booktitle = {International Conference on Machine Learning},
  year      = {1990},
  pages     = {304-312},
  doi       = {10.1016/B978-1-55860-141-3.50040-7},
  url       = {https://mlanthology.org/icml/1990/minton1990icml-issues/}
}