In deep learning, "softmax" is a commonly used activation function, which is mainly used in the output layer in multi-category classification tasks. "Softmax" is the abbreviation of "soft maximum", which is used to represent the probability of the corresponding class by converting the input real vector into a probability distribution.
Usage of softmax activation function in deep learning
Guess you like
Origin blog.csdn.net/qq_45560230/article/details/132831508
Recommended
Ranking