Checklist for reporting ML-based science

About. The REFORMS checklist lists items that should be reported in a scientific study that uses machine learning (ML) methods. It is intended to accompany the paper or report that introduces an ML model: for instance, as an appendix or supplemental material. The checklist consists of 32 questions spread across 8 modules. For each item, either list the section name, section number, or page number in the paper where the item is reported, or justify why a given item is not filled out. Note that not all of these items need to be reported in the main text of the paper; they could be reported in an appendix or supplementary files.

关于《REFORMS清单》列出了在使用机器学习（ML）方法的科学研究中应报告的项目。它旨在附在介绍ML模型的论文或报告之后：例如，作为附录或补充材料。该清单由分布在8个模块中的32个问题组成。对于每个项目，要么列出论文中报告该项目的章节名称、章节编号或页码，要么说明为什么某个项目未填写。请注意，并非所有这些项目都需要在论文正文中报告；它们可以在附录或补充文件中报告。