.png){kind=icon}
Return
Ensemble mating strategy in evolutionary algorithms for multiobjective optimization
.png){kind=icon}
PRE
AI
DOI：10.1007/s12293-025-00458-2.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Elite-based mating selection and neighbor-based mating restriction are two popular strategies for selecting promising parents for offspring generation in multiobjective evolutionary algorithms (MOEAs). However, few studies have explored the combination of these distinct mating strategies to enhance the search efficiency of MOEAs. To address this issue, we propose an ensemble mating strategy-based multiobjective evolutionary algorithm (EMSEA) that incorporates both mating selection and mating restriction strategies for promising offspring generation. In EMSEA, a clustering learning method is adopted to extract the population structure by partitioning the population into distinct clusters. To generate a new trial solution from the solutions within a cluster, we utilize a competitive approach to select an elite individual from the cluster as a mating parent, while restricting the selection of additional parents to the same cluster. We empirically compare the performance of our algorithm with several representative MOEAs on test instances with complex Pareto sets and Pareto fronts. The experimental results validate the effectiveness of the proposed ensemble mating strategy, as our algorithm outperforms the compared algorithms on these test instances.
Keywords:
Mating strategy
Offspring generation
Multi-objective optimization
Evolutionary algorithm
Journal
IF:
2.3
Papers:
458
Citations:
718
Organization
Cited Papers
.png){kind=icon}
ShareSaveShareSaveShareSave
View more