知识卡片

普通读书笔记卡 · 8966 · 生成模型:GPT-6

弱相容边稀释远处证据

理解等级:中级

示例中x2强烈偏好观测0,x1与x2只有轻微相同偏好,x3则偏好与x2相反。树上汇集给出x2强偏0、x1轻偏0、x3偏1,说明证据传播强度取决于相容势,不能认为一个强观测会让所有节点同样确定。

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

      For this simple toy example, we can tell what results to expect by inspection, then verify that BP is doing the right thing. Node x2 wants very much to look like y2 = 0, because ψ42(x2, y2) contributes a large valued to the posterior probability when x2 = y2 = 1 or when x2 = y2 = 0. From ψ12(x1, x2) we see that x1

    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:摘录中关于弱相容边稀释远处证据的定义、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:For this simple toy example, we can tell what results to expect by inspection, then verify that BP is doing the right thing. Node x2 wants very much to look like y2 = 0, because ψ42(x2, y2) contributes a large valued to the posterior probability when x2 = y2 = 1 or when x2 = y2 = 0. From ψ12(x1, x2) we see that x1