I'm studying classification and regression. But I met some barriers that make me tired.
The first barrier is this: I heard that the sigmoid function is used for binary classification and the softmax function is used for multi-class classification. I think the two functions are similar because their graphs are almost similar. But what features do make the sigmoid function used for binary classification and the softmax function used for multi-class classification?
The last barrier is this: In Logistic regression, when sigmoid and softmax functions are used in each situation?