Decision Trees are predictive models that map observations about an item to conclusions about the item's target value. They use a tree-like structure of decisions and their possible consequences, creating a powerful tool in Machine Learning for Classification and regression tasks.