.png){kind=icon}
.png){kind=icon}
Return
Population-level center-based sampling for meta-heuristic algorithms
.png){kind=icon}
OA
AI
DOI：10.1016/j.swevo.2024.101827.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In recent years, the challenge of enhancing the efficiency and effectiveness of meta-heuristic algorithms has gained significant attention. Center-based sampling has shown promise in addressing this challenge, yet its application often requires customization for specific algorithms, limiting its generalizability. This study identifies a gap in the literature regarding the operation-independent application of center-based sampling. To address this, we propose a novel center-based sampling strategy at the population level, which can be seamlessly integrated into any population-based optimization algorithm. Our approach employs a collaborative multi-parent method to generate multiple center-based solutions, thereby increasing diversity and exploiting the solution space more effectively. We introduce two specific strategies: cluster-driven center-based sampling for single-objective optimization and ranking-driven center-based sampling for multi-objective optimization. The performance of these strategies is evaluated using the benchmark functions for the CEC-2017 competition on 5 single- and 6 many-objective evolutionary algorithms, demonstrating 40% similar to 100% statistical fitness improvement ratio over parent meta-heuristic algorithms, respectively. These findings highlight the potential of population-level center-based sampling to enhance the performance of meta-heuristic algorithms.
Keywords:
Center-based sampling
Population-based algorithms
Single-objective optimization
Multi-objective optimization
Meta-heuristic algorithms
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
8.5
Papers:
2.2K
Citations:
1.0W