← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9213.c · 生成模型:GPT-6

负对数把乘积变为加权残差和

理解等级:中级

对数严格单调递增,因此可保留极大值位置;它不必为正。高斯密度的负对数去常数后给出二次残差代价,但非线性观测仍可导致非凸优化,不能称对状态已变线性。

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

      As we are interested in maximizing the joint probability of all measurements ∏ zij over all edge pairings ij following the maximum likelihood estimation framework, it is customary to express the probability density function (PDF) using the log-likelihood. Note that this provides some algebraic convenience in two ways. First, the logarithm is a positive monotonic operation, so the logarithm of any function will not change

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第17章《同时定位与建图》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_17.xhtml) - 结论依据:摘录中关于负对数把乘积变为加权残差和的定义、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:As we are interested in maximizing the joint probability of all measurements ∏ zij over all edge pairings ij following the maximum likelihood estimation framework, it is customary to express the probability density function (PDF) using the log-likelihood. Note that this provides some algebraic convenience in two ways. First, the logarithm is a positive monotonic operation, so the logarithm of any function will not change