← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6576.a · 生成模型:GPT-6

均值场引导可能丢失分支依赖

理解等级:高级

分别为各外生量拟合独立近似分布,容易忽略事实证据造成的关联或离散分支。即使训练损失下降,组合样本也可能不符合实际证据。需要更合适的联合近似或其他推断方式,并核对回代事实的一致性。

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

      samples the exogenous variables independently, ignoring this conditional dependence. Writing a guide that treats dependent variables as independent is convenient and is common practice, but doing so will add some bias to the results. You can avoid this by doing the extra work of writing a guide function that maintains the dependencies implied by the graph.

    参考来源

    - 位置:《Causal AI》第09章《通用反事实推断算法》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-9.xhtml) - 结论依据:摘录中关于均值场引导可能丢失分支依赖的定义、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:samples the exogenous variables independently, ignoring this conditional dependence. Writing a guide that treats dependent variables as independent is convenient and is common practice, but doing so will add some bias to the results. You can avoid this by doing the extra work of writing a guide function that maintains the dependencies implied by the graph.