How is supervised learning different from heuristics?

How is supervised learning different from heuristics?

• A. Supervised learning learns from historical examples, while heuristics rely on human-defined rules.
• B. Supervised learning uses trial-and-error; heuristics rely on labels.
• C. Supervised learning cannot learn from data; heuristics can.
• D. Supervised learning is only for images; heuristics only for text.

Answer: (A)

The key idea is how models acquire knowledge: supervised learning builds a model by using historical labeled examples to learn a function that maps inputs to outputs, and then applies that function to new data. Heuristics are fixed, human-defined rules or guidelines that don’t learn from data on their own. So the statement that supervised learning learns from historical examples while heuristics rely on human-defined rules best captures the difference. The other ideas don’t fit because supervised learning does learn from data (not just trial-and-error), heuristics don’t depend on labels in the same data-driven sense, and supervised learning isn’t limited to one data type while heuristics aren’t restricted to a single modality.

The key idea is how models acquire knowledge: supervised learning builds a model by using historical labeled examples to learn a function that maps inputs to outputs, and then applies that function to new data. Heuristics are fixed, human-defined rules or guidelines that don’t learn from data on their own. So the statement that supervised learning learns from historical examples while heuristics rely on human-defined rules best captures the difference. The other ideas don’t fit because supervised learning does learn from data (not just trial-and-error), heuristics don’t depend on labels in the same data-driven sense, and supervised learning isn’t limited to one data type while heuristics aren’t restricted to a single modality.