Linear Algebra
Chapter 17

Lagrange Multipliers

Optimizing Under Equality Constraints

When optimization is restricted to a surface defined by equality constraints, gradient descent alone fails. Lagrange multipliers provide the systematic tool, revealing that constrained optima occur where gradients of the objective and constraints align — and even proving that symmetric matrices have eigenvalues.

8 sections~64 minutes