知识卡片

普通读书笔记卡 · 8984.d · 生成模型:GPT-6

高斯极大似然用总体分母

理解等级:中级

对独立样本,一维高斯最大似然估计为μ̂=样本均值、v̂=N⁻¹Σ(xi−μ̂)²。这里方差分母是N,不是无偏估计常用的N−1。若样本全相同,最优方差逼近零、普通正方差密度出现退化,需约束或正则处理。

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

      You might recognize the solutions for and as the empirical mean and variance of the data, respectively. This makes sense: we have just shown that to maximize the probability of the data under a Gaussian, we should set the mean and variance of the Gaussian to be the empirical mean and variance of the data.

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第32章《生成模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c032.xhtml) - 结论依据:摘录中关于高斯极大似然用总体分母的定义、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:You might recognize the solutions for and as the empirical mean and variance of the data, respectively. This makes sense: we have just shown that to maximize the probability of the data under a Gaussian, we should set the mean and variance of the Gaussian to be the empirical mean and variance of the data.