· Popular understanding of multi-label classification and multi-classification

I think of a very appropriate example, which is similar to the question of multiple choice and single choice. Multiple classification (single choice) is to select the option with the maximum probability of being correct. Multiple labels (multiple choice) need to judge whether each option is correct And multi-label classification is transformed into multiple dichotomies by Singmod. The difficulty of multi-label classification is similar to the difficulty of multiple choice for single choice