AI engineering glossary · Calculus optimisation
What is KKT conditions?
KKT conditions (also called Karush-Kuhn-Tucker conditions): The Karush-Kuhn-Tucker conditions, necessary requirements for optimality in problems with inequality constraints: stationarity of the Lagrangian, feasibility, non-negative multipliers for inequalities, and complementary slackness. They are also sufficient for convex problems and underlie support vector machines.
KKT conditions: an example
For minimising x^2 subject to x >= 1, the multiplier is 2 at x = 1 with the constraint active; complementary slackness says the multiplier times the constraint slack is zero.
Related terms
More Calculus optimisation terms
- L-BFGS
- L-smoothness
- Lagrange multiplier
- Lagrangian
- Learning rate
- Learning rate schedule
- Learning rate warmup
- Level set
Where you learn it
KKT conditions is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.