Scholarlink

Distributed algorithm design and execution has been a central aspect of evolutionary computation research ever since. Be it to spread and balance workload over many computational nodes to improve runtime performance, or to facilitate new algorithmic strategies, such as co-evolutionary approaches, for instance, to optimize convergence behavior. Although many software-research projects support distributed evolutionary computation, we see potential to further improve the usability of respective algorithm design. In this paper, we present a containerized software system, incorporating a domain specific design language, to address this issue. The performed experiments prove the developed system’s applicability and support further research leads.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Composable Evolutionary Computation

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jan Zenisek,
  • Florian Bachinger,
  • Florian Holzinger,
  • Erik Pitzer,
  • Stefan Wagner,
  • Michael Affenzeller

摘要

Distributed algorithm design and execution has been a central aspect of evolutionary computation research ever since. Be it to spread and balance workload over many computational nodes to improve runtime performance, or to facilitate new algorithmic strategies, such as co-evolutionary approaches, for instance, to optimize convergence behavior. Although many software-research projects support distributed evolutionary computation, we see potential to further improve the usability of respective algorithm design. In this paper, we present a containerized software system, incorporating a domain specific design language, to address this issue. The performed experiments prove the developed system’s applicability and support further research leads.