arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.04831math.CTcs.PL
收藏

基于Catlab和SemanticModels的组合科学计算

Compositional Scientific Computing with Catlab and SemanticModels

Micah Halter, Evan Patterson, Andrew Baas, James Fairbanks

摘要:

科学计算目前是通过编写特定领域的建模框架来解决特殊类别的数学问题。由于应用范畴论为描述和分析数学的不同领域提供了抽象推理机制,因此它是为科学计算构建通用和可重用软件组件的天然平台。我们提出了Catlab.jl,它为这个项目提供了分类理论基础设施,以及SemanticModels.jl,它利用这个基础设施来完成特定的建模任务。这种方法通过将关联系统的数学建模应用于cospan代数来增强和自动化科学计算工作流程。

英文摘要:

Scientific computing is currently performed by writing domain specific modeling frameworks for solving special classes of mathematical problems. Since applied category theory provides abstract reasoning machinery for describing and analyzing diverse areas of math, it is a natural platform for building generic and reusable software components for scientific computing. We present Catlab.jl, which provides the category-theoretic infrastructure for this project, together with SemanticModels.jl, which leverages this infrastructure for particular modeling tasks. This approach enhances and automates scientific computing workflows by applying recent advances in mathematical modeling of interconnected systems as cospan algebras.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑