← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8967 · 生成模型:GPT-6

图的树宽影响精确代价

理解等级:高级

一般图的精确推断复杂度与消元时形成的团及树宽有关,而不只看节点总数或有没有环。低树宽的有环图仍可能高效精确求解,二维大网格则通常产生昂贵中间因子。把多个变量合为节点虽然可改变表示,却可能让状态空间指数扩大。

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

      The BP message update rules only work to give the exact marginals when the topology of the network is that of a tree or a chain. In general, one can show that exact computation of marginal probabilities for graphs with loops depends on a graph-theoretic quantity known as the treewidth of the graph [274]. For many graphical models of interest in vision, such as 2D

    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:摘录中关于图的树宽影响精确代价的定义、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:The BP message update rules only work to give the exact marginals when the topology of the network is that of a tree or a chain. In general, one can show that exact computation of marginal probabilities for graphs with loops depends on a graph-theoretic quantity known as the treewidth of the graph [274]. For many graphical models of interest in vision, such as 2D