← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6543.b · 生成模型:GPT-6

因果抽象要求保留相关关系

理解等级:中级

将像素等低层信息映射成高层变量,需要说明该映射如何保留目标因果问题的含义。普通特征工程以预测有用为目标,因果抽象还要求相应因果关系或干预语义保持适用;高层名称直观不代表映射已满足这些条件。

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

      causal abstraction. In machine learning, the term “feature engineering” applies to the task of computing useful high-level features from lower-level features. Causal abstraction differs in that requirement for causal rigor. You’ll find some sources for causal abstraction information in the book’s notes at https://www.altdeep.ai/p/causalaibook.

    参考来源

    - 位置:《Causal AI》第03章《构建因果图模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-3.xhtml) - 结论依据:摘录中关于因果抽象要求保留相关关系的定义、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:causal abstraction. In machine learning, the term “feature engineering” applies to the task of computing useful high-level features from lower-level features. Causal abstraction differs in that requirement for causal rigor. You’ll find some sources for causal abstraction information in the book’s notes at https://www.altdeep.ai/p/causalaibook.