Cross Entropy Error Function
二分类
L = 1 N ∑ i L i = 1 N ∑ i − [ y i l o g ( p i ) ] + ( 1 − y i ) l o g ( 1 − l o g ( p i ) ) ] L = \frac{1}{N}\sum_iL_i = \frac{1}{N}\sum_i-[y_ilog(p_i)]+(1-y_i)log(1-log(p_i))] L=N1∑iLi=N1∑i−[yilog(pi)]+(1−yi)log(1−log(pi))]
多分类
L = 1 N ∑ i L i = 1 N ∑ i − ∑ c = 1 m y i c l o g ( p i c ) L=\frac{1}{N}\sum_iL_i=\frac{1}{N}\sum_i -\sum_{c=1}^m y_{ic} log(p_{ic}) L=N1∑iLi=N1∑i−∑c=1myiclog(pic)
交叉熵损失函数及其与熵和KL散度的关系
最小化交叉熵等价于最小化KL散度等价于最大化对数似然估计。