知识卡片
一个因果模型定义多种分布
完整指定模型及干预算子后,可定义不同do条件下的一族分布。能在模型中定义或算出这些分布,不代表它们都已由观察数据识别;多个相容模型可能给出不同干预答案。定义、识别和数值估计属于不同步骤。
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
- · 干预与因果效应
原书出处 · 1 处
- · 查看原书依据
7.9 Suppose our DGP has variables X, Y, and Z. A traditional generative model (left) uses observations of X, Y, and Z to statistically learn a representation of P(X, Y, Z). A generative causal model (right) encodes a representation P(X, Y, Z) and distributions derived by interventions on X, Y, and Z. In that way, the generative causal model represents a broad family of distributions.
参考来源
- 位置:《Causal AI》第07章《干预与因果效应》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-7.xhtml)
- 结论依据:摘录中关于一个因果模型定义多种分布的定义、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:7.9 Suppose our DGP has variables X, Y, and Z. A traditional generative model (left) uses observations of X, Y, and Z to statistically learn a representation of P(X, Y, Z). A generative causal model (right) encodes a representation P(X, Y, Z) and distributions derived by interventions on X, Y, and Z. In that way, the generative causal model represents a broad family of distributions.