pyRMV：面向计算科学的可复用跨模型验证框架
pyRMV: Reusable, Cross-model Validation for Computational Science
浏览论文内容
中文总结 AI 辅助
pyRMV是一种与模型无关的计算模型验证方法，配套Python库含小鼠初级视觉皮层模型的验证项，可推广至多类模型，为模型验证标准化迈出重要一步。
中文摘要 AI 辅助
我们提出了一种与模型无关的计算模型验证方法，以及一个包含针对小鼠初级视觉皮层模型的多组验证项的Python库。该方法将模型视为预测系统属性的生成器而非目标系统的替代物，这使得验证项可轻松推广至多类模型，同时也让验证项的编写相对简便。文中还概述了该方法的若干局限性及可能的解决途径，这为解决我们认为计算科学中最重要的问题——模型验证的标准化——迈出了重要一步。
英文摘要
We present a model-agnostic approach to the validation of computational models, and a python library containing a suite of validations for mouse primary visual cortex models. By viewing the model not as a stand-in for the target system, but instead as a generator of predicted system properties, this approach allows validations to easily generalize to many different models, while also making them comparatively simple to write. Several limitations of this approach are outlined alongside possible means of addressing them. This presents an important step in addressing what we consider the most important problem in computational science: the standardization of model validation.
