← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9378.b · 生成模型:GPT-6

逐元素乘法与矩阵乘法不同

理解等级:中级

NumPy的*通常按逐元素与广播规则相乘,@或matmul执行矩阵乘法。逐元素并不要求形状完全相同,兼容广播也可发生;广播能运行却可能静默产生错误语义,需检查每个轴。

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

      But what if you just want to multiply each element (value) in the matrix by the corresponding element in another matrix? This is called element-wise multiplication, and you can use the normal multiplication star (\*) symbol you are used to. Notice that element-wise multiplication only works on matrices that have the exact same shape:

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》附录C《附录C:向量与线性代数》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/appendix-c.html) - 结论依据:摘录中关于逐元素乘法与矩阵乘法不同的定义、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:But what if you just want to multiply each element (value) in the matrix by the corresponding element in another matrix? This is called element-wise multiplication, and you can use the normal multiplication star (\*) symbol you are used to. Notice that element-wise multiplication only works on matrices that have the exact same shape: