Heuristic

A practical method that seeks useful results without guaranteeing an exact solution or optimality. Heuristics trade certainty or exhaustive search for speed, simplicity, or workable answers.

Greedy algorithm: An algorithm that repeatedly chooses the locally best available option. It illustrates a heuristic strategy that can find quick solutions without ensuring a global optimum.
Search algorithm: A procedure that explores possible states or solutions to find one meeting a goal. Heuristics often guide which parts of a search space to explore first.
Tversky–Kahneman heuristics: Psychological shortcuts such as availability and representativeness that influence judgments under uncertainty. These named mental shortcuts show how heuristic judgment operates beyond algorithms.
Exact algorithm: An algorithm that returns a correct solution satisfying its specification for every valid input. It contrasts practical shortcuts with methods that guarantee correctness under stated assumptions.
Hill climbing: A local-search method that repeatedly moves to a neighboring state with a better value. It uses local improvement as a practical guide, but can get stuck at a local optimum.
Optimization: The process of finding the best feasible choice under specified objectives and constraints. Many heuristics seek good, though not necessarily best, solutions to optimization problems.
Availability heuristic: A judgment shortcut that estimates likelihood by how readily examples come to mind. It is a specific cognitive heuristic that can speed judgment while distorting risk estimates.
Exhaustive search: A method that checks every candidate in a finite search space. Heuristics avoid the complete enumeration that exhaustive search requires.
Beam search: A search algorithm that retains only a fixed number of promising candidates at each step. It makes large searches manageable by discarding candidates that appear less promising.
Objective function: A function that assigns a numerical value to a candidate solution for evaluation. A heuristic needs a criterion, explicit or implicit, for judging candidate results.

Linked from 10 pages