A case study in the meta-reasoning procedure ND
Authors:
James J. Lu;
Jeffrey S. Rosenthal a;
Andrew E. shaffer b
| Affiliations: | a Department of Statistics, University of Toronto, 100 St. George Street, Toronto, Ontario, Canada M5S 3G3. |
| b Cornell University, Ithaca, NY 14853-2801, USA. |
DOI:
10.1080/0952813021000028621
Publication Frequency:
4 issues per year
Published in:
Journal of Experimental & Theoretical Artificial Intelligence,
Volume
15,
Issue
1
2003
, pages 47
- 71
Subjects:
Cognitive Artificial Intelligence.;
Cognitive Psychology;
Cognitive Science;
Evolutionary Computing;
Human Computer Intelligence;
Machine Learning - Design;
Neural Networks;
Robotics;
Systems & Controls;
Formats available:
PDF
(English)
View Article:
View Article (PDF)
Abstract
A new technique for improving the efficiency of propositional reasoning procedures is presented. The meta-search procedure, ND, is parameterized by a search procedure P and a real number for controlling the way in which P is applied to the given problem. Experiments using SATO on the domain of Crossword Puzzle Construction (CPC) illustrate the potential for ND. Variations of and future experiments with ND are discussed.
|
| Keywords: Problem Solving; Knowledge Representation; Search; Threshold Phenomena |

Download Citation

CiteULike
Del.icio.us
BibSonomy
Connotea