Modern heuristic techniques for combinatorial problems
Modern heuristic techniques for combinatorial problems
Approximation algorithms for bin packing: a survey
Approximation algorithms for NP-hard problems
Automatic Generation of Control Parameters for the Threshold Accepting Algorithm
MICAI '02 Proceedings of the Second Mexican International Conference on Artificial Intelligence: Advances in Artificial Intelligence
Learning the Empirical Hardness of Optimization Problems: The Case of Combinatorial Auctions
CP '02 Proceedings of the 8th International Conference on Principles and Practice of Constraint Programming
Learning dynamic algorithm portfolios
Annals of Mathematics and Artificial Intelligence
| Hi-index | 0.00 |
Some Hybrid Packing Systems integrate several algorithms to solve the bin packing problem (BPP) based on their past performance and the problem characterization. These systems relate BPP characteristics with the performance of the set of solution algorithms and allow us to estimate which algorithm is to yield the best performance for a previously unseen instance. The present paper focuses on the characterization of NP-hard problems. In related work, characterization metrics are traditionally oriented towards problem structure. In this work, we propose metrics based on descriptive statistics for the Bin Packing Problem (BPP). The proposed metrics are of general purpose, meaning that the metrics do not depend on problem structure and can be applied to BPP and other problems to complement existent metrics. The "enhanced" Hybrid Packing System outperforms the version that does not take advantage of the general-purpose metrics; the results obtained show a 3%-improvement with respect to the reference Packing System.