mathweb
Institutional Group • 13 people
Files / PhD Comprehensive Examination on Optimization

1. About this exam

Syllabus: Math 650

Background Topics:

  • Calculus: Taylor’s theorem and differential calculus of several variables
  • Basic results of analysis: compactness, connectedness, Heine-Borel, Bolzano-Weierstrass, Weierstrass, intermediate-value theorem, etc.
  • Symmetric matrices: diagonalization, and recognizing positive (semi)definite matrices (with eigenvalue test, Sylvester’s theorem, and Descartes’s rule of sign test)
Exam Topics:
  • Unconstrained optimization: first- and second-order tests for a local minimum, maximum, or saddle points of smooth functions of several variables, coercive functions and existence of global optimizers
  • Basics of convexity: convex sets, convex functions, Caratheodory’s theorem, Jensen’s inequality, firstand second-order characterization of convex functions, global optimality of a local minimizer of a convex function, variational inequality principle for minimization of a function on a convex set
  • Separation of convex sets: the basic separation theorem on separation of two convex sets, support plane theorem, familiarity with statement of proper separation theorem
  • Linear equality/inequality systems: Farkas’s lemma and Motzkin’s transposition theorem, linear programming duality
  • First-order optimality conditions in nonlinear programming: Fritz John’s theorem, Kuhn-Tucker conditions, Lagrange multipliers, constraint qualifications (Slater’s condition, Mangasarian-Fromovitz condition, etc.)
  • Second-order (necessary and sufficient) optimality conditions in nonlinear programming
  • Ability to solve concrete nonlinear programs through the use of the methods noted in the two preceding bullets
  • Duality: concept of the primal and dual programs with respect to a Lagrangian function, general correspondence between saddle points and duality, dual of a nonlinear program, familiarity with strong duality theorem for convex programming and the ability to use it in concrete problems, duality in quadratic programming
Suggested textbooks:
  1. Guler, Foundations of Optimization, 2010.
  2. Bertsekas, Nonlinear Programming, 2nd Edition, 1999.
  3. Borwein and Lewis, Convex Analysis and Nonlinear Optimization: Theory and Examples, 2009.
  4. Luenberger and Ye,  Linear and Nonlinear Programming, 3rd Edition, 2009.
  5. Bazaraa, Sherali and Shetty, Nonlinear Programming: Theory and Algorithms, 2nd Edition, 1993.