← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6586.b · 生成模型:GPT-6

界限应先检验极端分布

理解等级:高级

抄录识别界限前可检查上下界顺序、概率范围及退化分布。原书ATE界限图的上界在E恒0且I恒0、随机分配独立的例子中给出−1,而其下界为0,故不能作为通用有效式直接使用。应回到推导或原始文献核对符号。

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

      the size of your data increases. (Successful convergence assumes typical “regularity” conditions for Bayesian estimation. Results will depend on the quality of the prior.) But even if it doesn’t converge to the true value, your assumptions might still enable convergence to a ballpark region around the true value that is small enough to be useful (this is called partial identification, as described in section 10.9).

    参考来源

    - 位置:《Causal AI》第10章《识别与因果层级》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-10.xhtml) - 结论依据:摘录中关于界限应先检验极端分布的定义、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:the size of your data increases. (Successful convergence assumes typical “regularity” conditions for Bayesian estimation. Results will depend on the quality of the prior.) But even if it doesn’t converge to the true value, your assumptions might still enable convergence to a ballpark region around the true value that is small enough to be useful (this is called partial identification, as described in section 10.9).