OR 649 Nonlinear and Dynamic Programming
              Text:  Basic Optimization Methods by Brian Bunday.  (Additional references) 
  
            Course Outline:

                         1. Introduction

                         2. Mathematical Preliminaries
                                  A. Basic concepts and definitions
                                  B. Convex/Concave funtions
                                  C. Quadratic forms
                                  D. Brief review of linear algebra

                         3. Unconstrained Optimization
                                  A. Gradient search
                                  B. Fibonacci search
                                  C. Golden Section search
                                  D. Polynomial interpolation
                                  E. Hooke-Jeeves method
                                  F. Davidon-Fletcher-Powell method
                                  G. Other approaches

                        4. Constrained Optimization
                                  A. Karush-Kuhn-Tucker conditions
                                  B. Saddlepoint theorem
                                  C. Quadratic programming
                                        a. Applications
                                        b. Wolfe's algorithm
                                        c. Using LINDO
                                  D. Linearly constrained problems
                                       a. Separable programming
                                       b. Convex Simplex method
                                       c. Reduced Gradient method
                                       d. Other approaches
                                  E. Nonlinearly constrained problems
                                       a. Separable programming
                                       b. SUMT
                                       c. Penalty methods
                                       d. Generalized Reduced Gradient method
 
                        5. Geometric Programming

                        6. Dynamic Programming
 

            Grading:  The final grade is based on the results of homework, tests, and a comprehensive final exam.
                             Item weights will be provided in class.

                    The grading scale is:  90-100% A   80-89% B   70-79% C   60-69% D   0-59% F.   
 

            Prerequisites:  GBA 661 or CS 440 or equivalent.
 
 

            Link to downloadable programs
 
            Homework page