sigmoid函数图像演变
（便于记忆）
g
(
s
)
=
e
s
g(s) = e^s
g(s)=es
g
(
s
)
=
e
−
s
g(s) = e^{-s}
g(s)=e−s
g
(
s
)
=
1
+
e
−
s
g(s) = 1+ e^{-s}
g(s)=1+e−s
g
(
s
)
=
1
1
+
e
−
s
g(s) = \frac{1}{1+ e^{-s}}
g(s)=1+e−s1
sigmoid函数导数推导
f ( x ) = 1 1 + e − x = e x 1 + e x = 1 − ( e x + 1 ) − 1 f(x) = \frac{1}{1+e^{-x}} = \frac{e^x}{1+e^x} = 1-(e^x+1)^{-1} f(x)=1+e−x1=1+exex=1−(ex+1)−1
f ′ ( x ) = ( − 1 ) ( − 1 ) ( 1 + e x ) − 2 e x = e x ( 1 + e x ) 2 = e x 1 + 2 e x + e 2 x f^{'}(x) = (-1)(-1)(1+e^{x})^{-2}e^{x}=\frac{e^x}{(1+e^x)^2}=\frac{e^x}{1+2e^x+e^{2x}} f′(x)=(−1)(−1)(1+ex)−2ex=(1+ex)2ex=1+2ex+e2xex
接下来把所有 e x e^x ex转换为 e − x e^{-x} e−x（为了最后把导数写成 f ( x ) 的 形 式 f(x)的形式 f(x)的形式
f ′ ( x ) = e − x e − 2 x + 2 e − x + 1 = e − x ( e − x + 1 ) 2 = 1 1 + e − x e − x 1 + e − x = f ( x ) ( 1 − f ( x ) ) f^{'}(x) = \frac{e^{-x}}{e^{-2x}+2e^{-x}+1} = \frac{e^{-x}}{(e^{-x}+1)^2}=\frac{1}{1+e^{-x}} \frac{e^{-x}}{1+e^{-x}}=f(x)(1-f(x)) f′(x)=e−2x+2e−x+1e−x=(e−x+1)2e−x=1+e−x11+e−xe−x=f(x)(1−f(x))
咱也不知道为啥也写成这种形式
2073
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
