← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6589.b · 生成模型:GPT-6

元学习器输出条件均值差

理解等级:中级

T-learner分别拟合处理组与对照组的结果均值,再在同一协变量值比较预测。输出是条件平均效应的估计,不能解释为已观察到该个体的两种潜在结果;稀少组、支持缺口和模型误设都可放大差值误差。

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

      speaking, meta learners train a model (or models) of the outcome given the treatment variable and the confounders, and then account for the difference in prediction across treatment and control values of the treatment variable. They are particularly focused on highlighting heterogeneity of treatment effects across the data. The following code shows a meta learner example called a T-learner that uses a random forest predictor.

    参考来源

    - 位置:《Causal AI》第11章《构建因果推断工作流》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-11.xhtml) - 结论依据:摘录中关于元学习器输出条件均值差的定义、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:speaking, meta learners train a model (or models) of the outcome given the treatment variable and the confounders, and then account for the difference in prediction across treatment and control values of the treatment variable. They are particularly focused on highlighting heterogeneity of treatment effects across the data. The following code shows a meta learner example called a T-learner that uses a random forest predictor.