Lecture 17.2: from the KKT conditions to duality
Aggregazione dei criteri
The easy case: optimality conditions for equality constraints, some strange multipliers \mu appear. The inequality constraints case: cone of first-order feasible directions, constraint qualifications. First-order optimality conditions: KKT (what complementary slackness means). A glimpse to second-order optimality conditions, the role of the Lagrangian function. From the Lagrangian function to the dual function: Lagrangian duality. Duality and convexity. Take away: duality is a crucial feature in constrained optimization.