the words behind a decision you can defend.
Each term here gets a definition in one sentence, an explanation in a paragraph, and a note on how kint uses it.
start here
five terms in every kint answer.
Each kint result has the same parts. These are the words to read first.
- 01objective functionWhat the plan is best at.
- 02binding constraintThe rule that decides the outcome.
- 03sensitivity analysisWhat one more unit of a resource is worth.
- 04optimality gapHow close to the best possible plan, with a bound.
- 05deterministic replayThe recorded plan, reproduced from pinned inputs.
Matches the term, its abbreviation and its definition. Esc clears the field.
B
- binding constraintA constraint that the optimal plan meets exactly. Loosen it and the result improves; loosen any non-binding constraint and nothing changes.
- blackbox optimizationOptimization of a system whose behaviour is known only through data or simulation, not through formulas you can write down.
- branch and bound B&BAn algorithm for integer and combinatorial problems that splits the search into subproblems and discards those that cannot contain a better solution.
C
- combinatorial optimizationFinding the best solution among a finite but enormous set of possibilities, where each solution is a combination of discrete choices.
- constraint programming CPA way to solve combinatorial problems by stating variables, their possible values and the rules that must hold together.
D
N
O
- objective functionThe quantity an optimization model maximises or minimises, such as total cost, distance, delay or margin.
- operations research ORThe discipline of applying mathematical modelling, optimization, simulation and statistics to better decisions.
- optimality gapThe distance between the best plan found and the best plan that could possibly exist, as proven by the solver's bound.
S
- scenario analysisSolving the same decision under several sets of assumptions and comparing the results.
- sensitivity analysisMeasuring how the optimal result changes when an input or a constraint changes.
- shadow priceThe change in the optimal objective value when the right-hand side of a constraint grows by one unit.
- simplex methodAn algorithm for linear programs that moves along the edges of the feasible region from corner to corner until no neighbour is better.
- solver certificateThe evidence a solver produces that a solution is optimal or within a stated gap, such as a matching primal and dual bound.
- surrogate modelA fast machine learning model trained to imitate a slow simulation or a system known only from data.
see the terms on your own decision.
A binding constraint means more when it is your own crane, budget or rest period. Bring one decision and we show you each part of the answer.
- Problem types
- routing, scheduling, allocation, stock, price
- Comparisons
- kint next to other approaches
- Short answers
- frequently asked questions
- Contact
- hello@kint.pro