Open junxnone opened 1 year ago
Sigmoid Function
Logistic Function
sigmoid
softmax
Softmax
Soft
max
Vector
概率 Vector
$softmax(\vec{z}){1}=\frac{e^{z{1}}}{e^{z{1}} + e^{z{2}}}=\frac{e^{x}}{e^{x} + e^{0}}=\frac{e^{x}}{e^{x} + 1}=\frac{1}{1+e^{-x} }$
Reference
Brief
Sigmoid Function
==Logistic Function
sigmoid
和softmax
等价- 多标签分类/Multi-Label
Softmax
Softmax
=Soft
+max
Vector
映射到概率 Vector
Examples
二分类时
sigmoid
和softmax
等价$softmax(\vec{z}){1}=\frac{e^{z{1}}}{e^{z{1}} + e^{z{2}}}=\frac{e^{x}}{e^{x} + e^{0}}=\frac{e^{x}}{e^{x} + 1}=\frac{1}{1+e^{-x} }$
对比