当前位置：网站首页>Binary cross entropy loss function

Binary cross entropy loss function

2022-04-22 02:27:00 【ManiacLook】

Cross entropy loss

Cross entropy ：
H(y⁽ⁱ⁾, $\hat{y}$ ⁽ⁱ⁾) = - $\sum_{j=0}^q$ y_j⁽ⁱ⁾ log $\hat{y}$ _j⁽ⁱ⁾

Where vector y⁽ⁱ⁾ Dimension is (1, q),q Is the number of features . Cross entropy only concerns the prediction probability of the correct category ,y⁽ⁱ⁾ Only one value in the vector is 1, The rest are all 0.

The cross entropy loss function is defined as ：

L = $\frac{1}{n}$ $\sum_{i = 1}^n$ H(y⁽ⁱ⁾, $\hat{y}$ ⁽ⁱ⁾)

Binary cross entropy loss

L(w) = - $\sum_{i=0}^N$ [y_ilog $\sigma$ (x_i) + (1 - y_i)log(1 - $\sigma$ (x_i))]

among , $\sigma$ (x) yes sigmoid function , namely ：

$\sigma$ (x) = $\frac{1}{1 + e^{-z}}$

explain

A visual way to explain logarithmic loss
Detailed explanation of cross entropy

版权声明
本文为[ManiacLook]所创，转载请带上原文链接，感谢
https://yzsam.com/2022/04/202204220222434913.html

当前位置：网站首页>Binary cross entropy loss function

Binary cross entropy loss function

Cross entropy loss

Binary cross entropy loss

explain

边栏推荐

猜你喜欢

随机推荐