optimeerimise

Optimeerimise, or optimization, is the systematic process of making a system, process, or decision operate at its best within defined constraints. It involves selecting the best feasible option to maximize or minimize a chosen objective, such as cost, time, or quality. The field spans mathematics, computer science, operations research, and engineering, and can be applied to physical, computational, and organizational systems.

Formal framework typically includes an objective function that expresses the goal to optimize, decision variables representing controllable choices, and constraints that limit feasible solutions. Problems may be deterministic and well-defined, or stochastic with uncertainty. The result is a solution or set of Pareto-optimal solutions in multi-objective cases.

Common methods include mathematical programming (linear, integer, nonlinear), convex optimization, dynamic programming, and gradient-based methods. When exact solutions are impractical, heuristics and metaheuristics (genetic algorithms, simulated annealing, tabu search) offer approximate but often useful results. Stochastic optimization handles uncertainty with probabilistic models.

Typical workflow: problem formulation, model construction, data collection, selecting an algorithm, solving, and validation. Solutions are deployed with monitoring to ensure sustained performance and updates when conditions change.

Applications span supply chain and logistics, production planning, network design, energy systems, scheduling, pricing, and machine learning hyperparameter tuning, among others.

Historically, optimization emerged in operations research and mathematical programming in the 20th century, with early work on linear programming by Dantzig and Kantorovich, and the calculus of variations preceding these developments. The Estonian term optimeerimise is tied to the broader aim of achieving best possible performance.