mitteeoptimaalne

Mitteeoptimaalne is a term used to describe something that does not achieve the optimum for a given objective function. In Estonian contexts it is applied to solutions, estimators, or decision rules that fall short of the best possible value according to the specified criterion. More common form is mitte-optimaalne, but mitteeoptimaalne appears in some texts as a variant spelling.

In optimization theory and related disciplines, mitteeoptimaalne lahendus refers to a solution whose objective value is not the global best. For minimization problems, a mitteeoptimaalne solution has a higher objective value than the global minimum; for maximization, a lower value than the global maximum. The gap between the non-optimal value and the optimum is called the suboptimality gap.

Suboptimality can arise for several reasons: use of heuristic or approximate algorithms, computational limits, model simplifications, noisy data, or incomplete information. Despite not achieving optimality, a mitteeoptimaalne solution may be preferable if it offers advantages in speed, scalability, robustness, or ease of implementation.

The concept is distinct from local optima: a solution can be locally optimal but globally non-optimal, i.e., mitteeoptimaalne with respect to the global criterion. Related terms include suboptimal, near-optimal, and approximation algorithms, each with its own nuances.

In practice, identifying suboptimality helps balance accuracy and resource use. See also the broader notions of optimality, global and local optima, and Pareto optimality, which provide formal frameworks for comparing solutions.