知识卡片

普通读书笔记卡 · 6548.c · 生成模型:GPT-6

祖先裁剪需检查全部路径

理解等级:高级

在选定的查询子图中,较近节点若截住更远祖先到目标的相关路径,有时可省去更远祖先。但若被省变量仍共同影响保留节点,必须表示诱导依赖或承认隐藏混杂,不能一般声称整个保留集合仍因果充分。

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

      Figure 3.17 Once you include a common cause, you don’t have to include any earlier common causes on the same paths to the step 1 variables. In figure 3.17, W2 is on W0’s path to Y and U, but we include W0 because it has its own path to V. In contrast, while C is a common cause of V, Y, and U, W0 is

    参考来源

    - 位置:《Causal AI》第03章《构建因果图模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-3.xhtml) - 结论依据:摘录中关于祖先裁剪需检查全部路径的定义、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:Figure 3.17 Once you include a common cause, you don’t have to include any earlier common causes on the same paths to the step 1 variables. In figure 3.17, W2 is on W0’s path to Y and U, but we include W0 because it has its own path to V. In contrast, while C is a common cause of V, Y, and U, W0 is