Course overview

STATS 606: Computation and Optimization Methods in Statistics

University of Michigan

including slides by Stephen Boyd and Lieven Vandenberghe

Course goals

1. modeling: recognize and formulate (statistical) problems as convex optimization problems

2. algorithms/methods: develop (pseudo)code for solving convex optimization problems

   1. 1st-order algorithms: (sub)gradient method, proximal methods, stochastic approximation etc

   2. decomposition/splitting methods: dual decomposition, multiplier methods (eg ADMM), operator splitting etc

3. theory: characterize optimal solutions, study their sensitivity to problem parameters etc