← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6552.k · 生成模型:GPT-6

离散化改变了被检验变量

理解等级:中级

把连续年龄转为年龄段便于分类检验,但新变量丢失区间内信息。尤其把条件变量粗化可能残留依赖,使原变量独立关系不再成立;这种简化需记录切点及对结论的影响,不能视为等价无损预处理。

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

      on building your DAG and using that DAG in downstream causal inferences. For example, if I had a mix of continuous and discrete variables, then rather than implementing a test that could accommodate my different data types, I would discretize my continuous variables (for example, turning age as time since birth into age brackets) and use a vanilla chi-squared test, to keep things moving along.

    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:摘录中关于离散化改变了被检验变量的定义、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:on building your DAG and using that DAG in downstream causal inferences. For example, if I had a mix of continuous and discrete variables, then rather than implementing a test that could accommodate my different data types, I would discretize my continuous variables (for example, turning age as time since birth into age brackets) and use a vanilla chi-squared test, to keep things moving along.