← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8752.b · 生成模型:GPT-6

测试策略先识别被测责任

理解等级:中级

先列数据处理、训练管线、服务和模型等对象,再选择直接检查其责任的测试。用整体准确率检查一个确定转换过于间接,用一个函数例子判断整体模型质量又不充分。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Component, or subject under test (SUT) Type of test What good looks like Software logic Logic, data transformations, feature engineering Unit tests Tests enumerate scenarios of how we interact with a function or class and specify the correct and expected behavior. Fast-running: Tens of tests run in seconds; even hundreds of tests can run within one to two minutes. Internal “private” functions may not need

    参考来源

    - 位置:《Effective Machine Learning Teams (David Tan, Ada Leung, David Colls)》第05章《软件自动化测试》(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/ch05.html) - 结论依据:摘录中关于测试策略先识别被测责任的定义、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:Component, or subject under test (SUT) Type of test What good looks like Software logic Logic, data transformations, feature engineering Unit tests Tests enumerate scenarios of how we interact with a function or class and specify the correct and expected behavior. Fast-running: Tens of tests run in seconds; even hundreds of tests can run within one to two minutes. Internal “private” functions may not need