A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Constrained optimization models engineering problems: maximizing profit subject to resource constraints, minimizing surface area for fixed volume, finding shortest paths on surfaces. Lagrange multipliers solve these systematically.
You learned Lagrange multipliers as a method for finding critical points of a function on a constraint surface. Constrained optimization applications ask: once you have that method, what real problems does it solve, and how do you set them up correctly?
The setup always has the same structure. There is an objective function f(x₁, …, xₙ) you want to maximize or minimize, and one or more constraint equations g(x₁, …, xₙ) = c that restrict which points are feasible. Classic examples: maximize the volume of a box (objective) with fixed total surface area (constraint); minimize the cost of a cylindrical can (objective) with fixed volume (constraint); find the point on a plane closest to the origin (objective) subject to the plane equation (constraint). The Lagrange condition ∇f = λ∇g is the same in every case — only the algebra changes.
The geometric intuition is worth carrying into applications: at a constrained optimum, the level sets of f are tangent to the constraint curve or surface. If they were not tangent — if they crossed — you could slide along the constraint and improve f, so you would not yet be at an optimum. Tangency means the two gradients point in the same direction, which is exactly ∇f = λ∇g.
The Lagrange multiplier λ itself carries important information that is easy to overlook. Mathematically, λ = df*/dc, where f* is the optimal value and c is the constraint bound. In words: λ tells you how much the optimal objective value changes per unit relaxation of the constraint. In the box problem, λ would tell you how much extra volume you gain per additional unit of surface area. In an economics problem, λ is the shadow price — the maximum you would be willing to pay for one more unit of the constrained resource. When you report a constrained optimization solution, reporting λ alongside the optimal point often provides the most actionable information.