Knapsack problems: algorithms and computer implementations
Knapsack problems: algorithms and computer implementations
Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
Two-phases Method and Branch and Bound Procedures to Solve the Bi–objective Knapsack Problem
Journal of Global Optimization
EMO'05 Proceedings of the Third international conference on Evolutionary Multi-Criterion Optimization
Initial population construction for convergence improvement of MOEAs
EMO'05 Proceedings of the Third international conference on Evolutionary Multi-Criterion Optimization
| Hi-index | 0.00 |
Population heuristics present native abilities for solving optimization problems with multiple objectives. The convergence to the efficient frontier is improved when the population contains `a good genetic information'. In the context of combinatorial optimization problems with two objectives, the supported solutions are used to elaborate such information, defining a resolution principle in two phases. First the supported efficient solution set, or an approximation, is computed. Second this information is used to improve the performance of a population heuristic during the generation of the efficient frontier. This principle has been experimented on two classes of problems : the 1 || (Σ,Ci; Tmax) permutation scheduling problems, and the biobjective 0-1 knapsack problems. The motivations of this principle are developed. The numerical experiments are reported and discussed.