Improved Bounds on the Complexity of kB-Consistency

Abstract

Abstract-consistencies form the class of strong consis-tencies used in interval constraint programming. We survey, prove, and give theoretical motivations to some technical improvements to a naive ¢¤£ consistency algorithm. Our contribution is twofold: on the one hand, we introduce an ¥ optimal-consistency algorithm whose time-complexity ¦¨§�©�������� of improves the known bound by a � factor is the number of constraints, � is the number of variables, and � is the maximal size of the intervals of the box). On the other hand, we prove that improved bounds on time complexity can effectively be reached for higher values of ¢. These results are obtained with very affordable overheads in terms of space complexity.

Cite

Text

Bordeaux et al. "Improved Bounds on the Complexity of kB-Consistency." International Joint Conference on Artificial Intelligence, 2001.

Markdown

[Bordeaux et al. "Improved Bounds on the Complexity of kB-Consistency." International Joint Conference on Artificial Intelligence, 2001.](https://mlanthology.org/ijcai/2001/bordeaux2001ijcai-improved/)

BibTeX

@inproceedings{bordeaux2001ijcai-improved,
  title     = {{Improved Bounds on the Complexity of kB-Consistency}},
  author    = {Bordeaux, Lucas and Monfroy, Éric and Benhamou, Frédéric},
  booktitle = {International Joint Conference on Artificial Intelligence},
  year      = {2001},
  pages     = {303-308},
  url       = {https://mlanthology.org/ijcai/2001/bordeaux2001ijcai-improved/}
}