sparse_categorical_crossentropy 和 categorical_crossentropy 有什么区别? [英] What is the difference between sparse_categorical_crossentropy and categorical_crossentropy?
问题描述
sparse_categorical_crossentropy
和 categorical_crossentropy
有什么区别?什么时候应该使用一种损失而不是另一种?例如,这些损失是否适合线性回归?
What is the difference between sparse_categorical_crossentropy
and categorical_crossentropy
? When should one loss be used as opposed to the other? For example, are these losses suitable for linear regression?
推荐答案
简单:
categorical_crossentropy
(cce
) 生成一个单热数组,其中包含每个类别的可能匹配项,sparse_categorical_crossentropy
(scce
) 生成最有可能匹配类别的类别索引.
categorical_crossentropy
(cce
) produces a one-hot array containing the probable match for each category,sparse_categorical_crossentropy
(scce
) produces a category index of the most likely matching category.
考虑一个有 5 个类别(或类)的分类问题.
Consider a classification problem with 5 categories (or classes).
在
cce
的情况下,one-hot target可能是[0, 1, 0, 0, 0]
,模型可以预测[.2, .5, .1, .1, .1]
(可能是对的)
In the case of
cce
, the one-hot target may be[0, 1, 0, 0, 0]
and the model may predict[.2, .5, .1, .1, .1]
(probably right)
在scce
的情况下,目标索引可能是[1],模型可能预测:[.5].
In the case of scce
, the target index may be [1] and the model may predict: [.5].
现在考虑一个有 3 个类的分类问题.
Consider now a classification problem with 3 classes.
- 在
cce
的情况下,one-hot 目标可能是[0, 0, 1]
并且模型可以预测[.5, .1, .4]
(可能不准确,因为它给第一类的概率更大) - 对于
scce
,目标索引可能是[0]
,模型可能预测[.5]
莉>
- In the case of
cce
, the one-hot target might be[0, 0, 1]
and the model may predict[.5, .1, .4]
(probably inaccurate, given that it gives more probability to the first class) - In the case of
scce
, the target index might be[0]
, and the model may predict[.5]
许多分类模型产生 scce
输出是因为您节省了空间,但会丢失很多信息(例如,在第二个示例中,索引 2 也非常接近.)我通常更喜欢 cce
输出模型可靠性.
Many categorical models produce scce
output because you save space, but lose A LOT of information (for example, in the 2nd example, index 2 was also very close.) I generally prefer cce
output for model reliability.
scce
有多种使用情况,包括:
There are a number of situations to use scce
, including:
- 当您的类是相互排斥的,即您根本不关心其他足够接近的预测时,
- 类别数量过多导致预测输出变得不堪重负.
这篇关于sparse_categorical_crossentropy 和 categorical_crossentropy 有什么区别?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!