知识卡片
有环消息传递属于近似
在有环图上初始化消息并反复使用局部BP更新,可能获得有用近似,但证据会沿环重复流动。算法可能振荡、不收敛,或收敛到不准确结果。需要设置迭代上限、收敛准则并检查结果,不能将“迭代到收敛”当作必然可执行的保证。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 概率图模型
原书出处 · 1 处
- · 查看原书依据
Message update rules are described locally, and one might imagine that it is a useful local operation to perform, even without the global guarantees of ordinary BP. It turns out that is true. Here is the algorithm, also know as loopy belief propagation algorithm:
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:摘录中关于有环消息传递属于近似的定义、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:Message update rules are described locally, and one might imagine that it is a useful local operation to perform, even without the global guarantees of ordinary BP. It turns out that is true. Here is the algorithm, also know as loopy belief propagation algorithm: