← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8851.b · 生成模型:GPT-6

正规方程求解需区分秩条件

理解等级:高级

平方损失的驻点满足XᵀXθ=Xᵀy。只有X列满秩时,XᵀX才可逆并得到唯一解θ*=(XᵀX)⁻¹Xᵀy;秩亏时可用伪逆选择最小范数解,最优参数也可能不唯一。原书称二次型只有一个极小值缺少这一条件。

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

      We set this derivative to zero and solve for θ*: (9.12) (9.13) (9.14) The θ* defines the best fitting line to our data, and this line can be used to predict the y value of future observations of x (figure 9.5). Figure 9.5: A best fit line is a visualization of a function fθ, that predicts the y-value for each input x-value. We can now

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第09章《学习导论》(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/c009b.xhtml) - 结论依据:摘录中关于正规方程求解需区分秩条件的定义、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:We set this derivative to zero and solve for θ*: (9.12) (9.13) (9.14) The θ* defines the best fitting line to our data, and this line can be used to predict the y value of future observations of x (figure 9.5). Figure 9.5: A best fit line is a visualization of a function fθ, that predicts the y-value for each input x-value. We can now