1. Matrix multiplication (vector inner product)
d=np.dot(a,a)#matrix multiplication
e=a@a#matrix multiplication
2. Hadamard product
b=a**2#Element level multiplication, equivalent to a*a, equivalent to np.multiply(a,a)
c=a*a#Element level multiplication, equivalent to a**2
f=np.multiply(a,a)#Element level multiplication, the shape can be different
demo:
import numpy as np
a = np.array([1,2])
print(f"a{
a}")
# >>>矩阵乘法（向量内积）
d=np.dot(a,a) #矩阵乘法
e=a@a #矩阵乘法
print(f"d {
d}")
print(f"e {
e}")
# >>>哈达玛(Hadamard)积
b=a**2 #元素级别相乘，等同于a*a,等同于np.multiply(a,a)
c=a*a #元素级别相乘，等同于a**2
f=np.multiply(a,a)#元素级别相乘,形状可以不一样
print(f"b{
b}")
print(f"c{
c}")
print(f"f{
f}")
aa =np.array([[1,2],[1,2]])
print(f"aa{
aa}")
g=a*aa
h=np.multiply(a,aa)
i=np.dot(a,aa)
j=a@aa
print(f"g{
g}")
print(f"h{
h}")
print(f"i{
i}")
print(f"j{
j}")
good post
Several common products in machine learning are introduced in more detail, you can directly see this