知识卡片
先验集中不能代替识别
贝叶斯后验可能因先验或参数限制而在不可识别方向上集中。应分别报告哪些约束来自数据、哪些来自图或函数假设、哪些来自先验;后验区间狭窄并不证明目标由指定观测分布唯一决定。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 识别与因果层级
原书出处 · 1 处
- · 查看原书依据
Bayesian network)—a probabilistic model trained on a causal DAG. A causal DAG by itself is a level 2 set of assumptions; assumptions about what causes what. Generally, assumptions that let you deduce the consequences of an intervention are level 2 assumptions.
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:摘录中关于先验集中不能代替识别的定义、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:Bayesian network)—a probabilistic model trained on a causal DAG. A causal DAG by itself is a level 2 set of assumptions; assumptions about what causes what. Generally, assumptions that let you deduce the consequences of an intervention are level 2 assumptions.