← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6551 · 生成模型:GPT-6

图分离与分布独立分层

理解等级:中级

d分离是由给定图和条件集决定的结构事实;条件独立是关于联合分布的事实。全局马尔可夫假设连接两者,使图上d分离推出分布独立;有限数据中的检验结果则是第三层,仍可能有统计误差。

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

      name is a misnomer; it does not “get independencies”; it gets d-separations. Again, don’t conflate d-separation in the causal graph with conditional independence in the joint probability distribution entailed by the DGP the graph is meant to model. Keeping this distinction in your mind will help you with the next task: using d-separation to test a DAG against evidence of conditional independence in the data.

    参考来源

    - 位置:《Causal AI》第04章《用因果约束检验DAG》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-4.xhtml) - 结论依据:摘录中关于图分离与分布独立分层的定义、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:name is a misnomer; it does not “get independencies”; it gets d-separations. Again, don’t conflate d-separation in the causal graph with conditional independence in the joint probability distribution entailed by the DGP the graph is meant to model. Keeping this distinction in your mind will help you with the next task: using d-separation to test a DAG against evidence of conditional independence in the data.