← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8963.b · 生成模型:GPT-6

平方损失选择后验均值

理解等级:中级

在平方误差及相应矩存在的条件下,最小后验期望损失的估计是E[x|y]。对向量,逐分量的后验均值构成整体均值,因此可先求每个节点的边缘分布再计算均值。若变量是无序类别,数值编码的平均未必是合法或有意义的类别。

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

      But we may want to penalize wrong answers as a function of how far they are from the correct answer. If that penalty function is the squared distance in x, then the point estimate that minimizes the average value of that error is called the minimum mean squared error estimate (MMSE). To find this estimate, we seek the that minimizes the squared error, weighted by

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第29章《概率图模型》(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/c029.xhtml) - 结论依据:摘录中关于平方损失选择后验均值的定义、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:But we may want to penalize wrong answers as a function of how far they are from the correct answer. If that penalty function is the squared distance in x, then the point estimate that minimizes the average value of that error is called the minimum mean squared error estimate (MMSE). To find this estimate, we seek the that minimizes the squared error, weighted by