Structured output is a type of Supervised Learning in which the output is not just a single value but a complex structured object, such as a tree, graph or sentence. Example: Predicting the structure of a protein molecule based on its chemical composition.