In your paper, on page 2, you said " In addition to the traditional cross-entropy loss. " Isn't that the cross-entropy loss includes the classification labels during computation? I don't really get it. Sorry for re-posting. But I don't think issue 1 is really closed.
In your paper, on page 2, you said " In addition to the traditional cross-entropy loss. " Isn't that the cross-entropy loss includes the classification labels during computation? I don't really get it. Sorry for re-posting. But I don't think issue 1 is really closed.
_Originally posted by @cherise215 in https://github.com/Jason-Qiu/Transfer_Knowledge_from_Language_to_Electrocardiography/issues/1#issuecomment-1557734177_