← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8763.a · 生成模型:GPT-6

方向性测试需要有依据的单调预期

理解等级:高级

保持其他条件,改变一个应影响结果的输入,并检查预测沿约定方向变化。方向应来自明确任务或约束而非未经证实相关性,也要限定有效区间,避免生成不可能样本再责怪模型。

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

      Directional expectation tests A directional expectation test is like an invariance test, except that we expect the prediction to change in a specific direction when we perturb the input data. In our example, we can write a test where all examples have the same attributes except in one dimension that is known to have an effect on the prediction, and assert that the model’s predicted

    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:摘录中关于方向性测试需要有依据的单调预期的定义、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:Directional expectation tests A directional expectation test is like an invariance test, except that we expect the prediction to change in a specific direction when we perturb the input data. In our example, we can write a test where all examples have the same attributes except in one dimension that is known to have an effect on the prediction, and assert that the model’s predicted