STAT 31000/CMSC 37811. Mathematical Computation II — Numerical Optimization

Department of Statistics
University of Chicago
Winter 2012

This course covers the fundamentals of continuous optimization, linear programming, and convex optimization. Students are expected to have a solid grounding in multivariate calculus and linear algebra, and will be expected to complete several substantial programming projects (using MATLAB) during the course.

The first part of this course will focus on techniques that form the basis for large scale optimization problems, namely, iterative methods for solving large sparse linear systems. We will emphasize connections to the second part of the course on nonlinear programming and to other courses in UChicago Optimization sequence:

http://optimization.uchicago.edu

We will discuss the following stationary methods:

Jacobi
Gauss-Seidel
Successive Over-Relaxation

the following semi-iterative methods:

Richardson
Steepest Descent
Chebyshev
Conjugate Gradient

and the following Krylov subspace methods:

CG
MINRES, SYMMLQ, LSQR
GMRES, QMR, BiCG

We will cover some basic ideas for preconditioning and stopping conditions.

This web page is for the first part of the course. The web page for the second part is here.

Announcements

02/10/12: Problem Set 4 posted.

02/02/12: Office hours tomorrow 3:00–3:30 and 4:30–5:00 on Fri, Feb 03.

01/29/12: Office hours next week will be 4:30–5:30 on Thu, Feb 02.

01/29/12: Problem Set 3 posted. Due 4:00 on Fri, Feb 03.

01/21/12: Problem Set 2 posted.

01/13/12: Problem Set 1 posted.

01/10/12: Lecture notes posted. Check your e-mail for the URL.

01/03/12: Check back regularly for announcements.

Lectures

Location: Eckhart Hall, Room 117

Times: 3:00–4:20pm.

Course staff

Instructors: Lek-Heng Lim (Part I) and Mihai Anitescu (Part II)
Office: Eckhart 122
lekheng(at)galton.uchicago.edu
Tel: (773) 702-4263
Office hours: Wed, 3:30–4:30pm.

Course Assistant: Yunda Zhong
Office: Ryerson N375
ydzhong(at)galton.uchicago.edu
Office hours: TBA

Problem Sets (Part I)

Problem set will be assigned weekly and will be due the following week. Collaborations are permitted but you will need to write up your own solutions.

Problem Set 4: PDF (posted: Feb 10; due: Feb 16)

Problem Set 3: PDF (posted: Jan 29; due: Feb 03)

Problem Set 2: PDF (posted: Jan 21; due: Jan 26)

Problem Set 1: PDF (posted: Jan 13; due: Jan 19)

Bug report on the problem sets or the solutions: lekheng(at)galton.uchicago.edu

Supplementary materials

Course homepage from Winter 2011.

Grades

Grade composition: No in-class examination. Grade based entirely on eight take-home problem sets.