Multi-Class vs. Multi-Label Classification | Notion

Multi-class Classification

One Label per Instance: In multi-class classification, each data instance must be assigned to one and only one class (label) out of multiple possible options.
Mutually Exclusive: The classes are mutually exclusive, meaning an instance cannot belong to more than one class at the same time.
Example: Categorizing images into different animal types: A picture can be either a "cat", a "dog", or a "bird", but not more than one at once.

Multi-label Classification

Multiple Labels per Instance: In multi-label classification, each instance can be associated with zero, one, or multiple labels simultaneously.
Not Mutually Exclusive: Classes are not mutually exclusive, allowing for the assignment of multiple relevant labels to a single example.
Example: Tagging an article with multiple topics: An article could be about "technology", "sports", and "politics" all at the same time.

Feature	Multi-class Classification	Multi-label Classification
Labels	One and only one label per instance	Zero, one, or multiple labels per instance
Classes	Mutually exclusive	Can overlap and co-occur
Example	Classifying an email as "spam", "work", or "personal"	Tagging a blog post as "Machine Learning", "NLP", and "Python"