知识卡片

普通读书笔记卡 · 6562.b · 生成模型:GPT-6

非高斯不自动保证结果非负

理解等级:中级

LiNGAM保留线性赋值但采用非高斯独立噪声,其可识别结论还需相应条件。把噪声改为正Gamma在书中正系数正偏移例可保证正输出,却不能让所有非高斯线性模型天然非负;输入和系数的支持同样重要。

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

      6.4.5 Linear non-Gaussian additive models Linear non-Gaussian additive models (LiNGAM) are useful when the Gaussian assumption on exogenous variables is not appropriate. In our example, the amount of protein Y cannot be negative, but that can easily occur in a linear model if β0, x, or nx have low values. LiNGAM models remedy this by allowing the exogenous variable to have a non-normal distribution. Listing

    参考来源

    - 位置:《Causal AI》第06章《结构因果模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-6.xhtml) - 结论依据:摘录中关于非高斯不自动保证结果非负的定义、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:6.4.5 Linear non-Gaussian additive models Linear non-Gaussian additive models (LiNGAM) are useful when the Gaussian assumption on exogenous variables is not appropriate. In our example, the amount of protein Y cannot be negative, but that can easily occur in a linear model if β0, x, or nx have low values. LiNGAM models remedy this by allowing the exogenous variable to have a non-normal distribution. Listing