Convex Analysis and Optimization

Lecture Notes

Complete set of lecture notes (PDF - 8.5MB)

LEC #

TOPICS

LECTURE NOTES

1

The role of convexity in optimization, duality theory, algorithms and duality

(PDF)

2

Convex sets and functions, epigraphs, closed convex functions, recognizing convex functions

(PDF)

3

Differentiable convex functions, convex and affine hulls, Caratheodory's theorem, relative interior

(PDF)

4

Algebra of relative interiors and closures, continuity of convex functions, closures of functions, recession cones and lineality space

(PDF)

5

Directions of recession of convex functions, local and global minima, existence of optimal solutions

(PDF)

6

Nonemptiness of closed set intersections, existence of optimal solutions, linear and quadratic programming, preservation of closure under linear transformation

(PDF)

7

Partial minimization, hyperplane separation, proper separation, nonvertical hyperplanes

(PDF)

8

Convex conjugate functions, conjugacy theorem, support functions

(PDF)

9

Min common/max crossing duality, weak duality, constrained optimization and minimax, strong duality

(PDF)

10

Min common/max crossing duality theorems, strong duality conditions, existence of dual optimal solutions, nonlinear Farkas' lemma

(PDF - 1.5MB)

11

Min common/max crossing theorem III, nonlinear Farkas' lemma/linear constraints, linear programming duality, convex programming duality

(PDF)

12

Convex programming duality, optimality conditions, mixtures of linear and convex constraints, existence of optimal primal solutions, Fenchel duality, conic duality

(PDF)

13

Subgradients, Fenchel inequality, sensitivity in constrained optimization, subdifferential calculus, optimality conditions

(PDF)

14

Min-max duality, existence of saddle points

(PDF)

15

Problem structures, conic programming

(PDF)

16

Conic programming, semidefinite programming, exact penalty functions, descent methods for convex/nondifferentiable optimization, steepest descent method

(PDF)

17

Subgradient methods, calculations of subgradients, convergence

(PDF)

18

Approximate subgradient methods, ε-subdifferential, ε-subgradient methods, incremental subgradient methods

(PDF)

19

Return to descent methods, fixing the convergence problem of steepest descent, ε-descent method, extended monotropic programming

(PDF)

20

Approximation methods, cutting plane methods, proximal minimization algorithm, proximal cutting plane algorithm, bundle methods

(PDF)

21

Generalized polyhedral approximations in convex optimization

(PDF - 2.9MB)

22

Review of Fenchel duality, review of proximal minimization, dual proximal minimization algorithm, augmented Lagrangian methods

(PDF)

23

Interior point methods, barrier method, conic programming cases, path following

(PDF)

24

Review and epilogue

(PDF - 1.8MB)

 

Assignments

This section contains homework assignments and the midterm exam.

Homework Assignments

ASSIGNMENTS

SOLUTIONS

Homework 1 (PDF)

(PDF)

Homework 2 (PDF)

(PDF)

Homework 3 (PDF)

(PDF)

Homework 4 (PDF)

(PDF)

Homework 5 (PDF)

(PDF)

 

Exam

Midterm exam with solutions (PDF)