知识卡片
数据检查显化输入前提
在训练或推理入口检验字段、类型和约束可提前发现不符合预期的数据。schema只覆盖契约一部分,PII扫描也可能漏罕见实体;不能由数据检查通过推断模型一定不偏或法律要求全满足。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 模型自动化测试
原书出处 · 1 处
- · 查看原书依据
Data contract tests Data contract tests verify that the data being passed between components of an ML solution—e.g., the output of a data pipeline into a training pipeline—match the expected schema in terms of attribute names, value types, and value constraints. Instead of waiting for an ML training cycle to blow up midway—or worse, fail silently—due to an unexpected data schema, we verify that the
参考来源
- 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第06章《模型自动化测试》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)/_epub-src/OEBPS/ch06.html)
- 结论依据:摘录中关于数据检查显化输入前提的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:Data contract tests Data contract tests verify that the data being passed between components of an ML solution—e.g., the output of a data pipeline into a training pipeline—match the expected schema in terms of attribute names, value types, and value constraints. Instead of waiting for an ML training cycle to blow up midway—or worse, fail silently—due to an unexpected data schema, we verify that the