知识卡片

普通读书笔记卡 · 6566.e · 生成模型:GPT-6

图工具同名do未必固定取值

理解等级:中级

本书所用pgmpy版本的图do执行删入边,概率网络do还修改相关因子,但接口未指定目标值。不能仅凭方法名就认定已实现do(X=x);应检查目标核是否确实是固定x的点质量。该行为属于书中版本,不推为当前API承诺。

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

      listing 7.11, graph surgery is implemented in the do method in the DAG class. The BayesianNetwork class, our default for building causal graphical models, also has a do method. Like the DAG method, it takes an intervention target. At the time of writing, the method does not take an intervention value and thus does not satisfy the second element of the definition of ideal intervention.

    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:摘录中关于图工具同名do未必固定取值的定义、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:listing 7.11, graph surgery is implemented in the do method in the DAG class. The BayesianNetwork class, our default for building causal graphical models, also has a do method. Like the DAG method, it takes an intervention target. At the time of writing, the method does not take an intervention value and thus does not satisfy the second element of the definition of ideal intervention.