A Support-Based Algorithm for the Bi-Objective Pareto Constraint
Abstract
Bi-Objective Combinatorial Optimization problems are ubiquitous in real-world applications and designing approaches to solve them efficiently is an important research area of Artificial Intelligence. In Constraint Programming, the recently introduced bi-objective Pareto constraint allows one to solve bi-objective combinatorial optimization problems exactly. Using this constraint, every non-dominated solution is collected in a single tree-search while pruning sub-trees that cannot lead to a non-dominated solution. This paper introduces a simpler and more efficient filtering algorithm for the bi-objective Pareto constraint. The efficiency of this algorithm is experimentally confirmed on classical bi-objective benchmarks.
Cite
Text
Hartert and Schaus. "A Support-Based Algorithm for the Bi-Objective Pareto Constraint." AAAI Conference on Artificial Intelligence, 2014. doi:10.1609/AAAI.V28I1.9119Markdown
[Hartert and Schaus. "A Support-Based Algorithm for the Bi-Objective Pareto Constraint." AAAI Conference on Artificial Intelligence, 2014.](https://mlanthology.org/aaai/2014/hartert2014aaai-support/) doi:10.1609/AAAI.V28I1.9119BibTeX
@inproceedings{hartert2014aaai-support,
title = {{A Support-Based Algorithm for the Bi-Objective Pareto Constraint}},
author = {Hartert, Renaud and Schaus, Pierre},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2014},
pages = {2674-2679},
doi = {10.1609/AAAI.V28I1.9119},
url = {https://mlanthology.org/aaai/2014/hartert2014aaai-support/}
}