|« Previous: Lecture 21 Summary||Next: Lecture 23 Summary »|
Several of the iterative algorithms so far have worked, conceptually at least, by turning the original linear-algebra problem into a minimization problem. It is natural to ask, then, whether we can use similar ideas to solve more general optimization problems, which will be the next major topic in 18.335.
A simple algorithm, to start with: successive line minimization (for unconstrained local optimization with continuous design parameters), which leads us directly to nonlinear steepest-descent and thence to nonlinear conjugate-gradient algorithms. The key point being that, near a local minimum of a smooth function, the objective is typically roughly quadratic (via Taylor expansion), and when that happens CG greatly accelerates convergence. (Mentioned Fletcher–Powell heuristic to help "reset" the search direction to the gradient if we are far from the minimum and convergence has stalled.)
Application of nonlinear CG to Hermitian eigenproblems by minimizing the Rayleigh quotient (this is convex, and furthermore we can use the Ritz vectors to shortcut both the conjugacy and the line minimization steps).