CME 338: Large-Scale Numerical Optimization

Research Professor Michael Saunders, Stanford University

The notes below are initially from last year's class. They will be updated as we proceed through the spring quarter. Hard copies of the updated notes will be handed out in class.

Review of matrix factorizations
Overview of optimization software
Suggested projects
Iterative methods for symmetric
Iterative methods for square and rectangular systems
The primal simplex method
LUSOL: A basis factorization package
Basis LU updates
Primal-dual interior methods
MINOS Part 1 – the reduced-gradient method
Augmented Lagrangians – BCL methods
MINOS Part 2 – LCL methods
SQP Methods
SNOPT examples