What is a heuristic?

What is a heuristic?

• A. A hand-crafted rule-of-thumb used to make decisions without learning from data.
• B. A statistical model learned from data.
• C. An algorithm that always guarantees optimal results.
• D. A method that converts unsupervised data into labels.

Answer: (A)

A heuristic is a simple, hand-crafted rule-of-thumb used to guide decisions without learning from data. It relies on domain knowledge or experience to produce quick, practical answers, which is particularly useful when data is scarce or time is limited. Because it isn’t learned from data, a heuristic won’t automatically adapt to new situations and doesn’t guarantee the best possible outcome. The other descriptions describe different ideas: a statistical model learned from data is a data-driven approach; an algorithm that always guarantees optimal results claims perfect optimality (which a heuristic does not); and a method that converts unsupervised data into labels refers to clustering or labeling tasks in unsupervised learning, not a rule-of-thumb for decision-making.

A heuristic is a simple, hand-crafted rule-of-thumb used to guide decisions without learning from data. It relies on domain knowledge or experience to produce quick, practical answers, which is particularly useful when data is scarce or time is limited. Because it isn’t learned from data, a heuristic won’t automatically adapt to new situations and doesn’t guarantee the best possible outcome. The other descriptions describe different ideas: a statistical model learned from data is a data-driven approach; an algorithm that always guarantees optimal results claims perfect optimality (which a heuristic does not); and a method that converts unsupervised data into labels refers to clustering or labeling tasks in unsupervised learning, not a rule-of-thumb for decision-making.