Metaheuristic

Metaheuristic is a high-level problem-solving framework that guides a search process toward good solutions when exact methods are computationally infeasible. Unlike exact algorithms that guarantee optimal solutions, metaheuristics trade optimality guarantees for scalability and flexibility, making them the method of choice for combinatorial optimization, scheduling, routing, and design problems where the solution space is vast and the landscape is non-convex.

The term was coined by Fred Glover in 1986 to describe methods that transcend ("meta") simple heuristics by incorporating learning, memory, or stochastic exploration. The major families include:

• Trajectory