lidya multisite > Computer Engineering Department > CENG 504

About

CENG 504

Optimization Methods

Unconstrained optimization; one-dimensional search, gradient search methods. Constrained optimization; linear and integer programming. Heuristics. Non-derivative methods (metaheuristics); simulated annealing, genetic algorithms, neural networks.

Class Hours

3 lecture hours

Prerequisite: Knowledge of calculus and elementary programming.

OBJECTIVE

To provide the fundamental concepts and mathematical tools in optimization theory.

LEARNING OUTCOMES

  1. Ability to formulate and solve engineering problems
  2. Ability to classify various optimization methods
  3. Ability to propose an optimization technique for a specific problem
  4. Ability to use optimization tools

TEXTBOOK

Chong, E. K. P., and Zak, S. H., An Introduction to Optimization, Fourth Edition, John Wiley & Sons, 2013.

Week Topic
1 Introduction to optimization
Math. Review
2 Vector spaces and matrices, linear transformations, orthogonal projections
3 Quadratic forms, calculus, conditions for local minimizers
Unconstrained optimization
4 One dimensional search – golden section search, Fibonacci search
5 One dimensional search – Newton’s method
6 Multi-dimensional gradient methods – steepest descent, Newton’s method
7 Conjugate direction methods
Constrained optimization
8 Linear programming
9 Linear programming – simplex algorithm
10
 Integer programming
11
 Heuristics for optimization
Non-derivative methods (Metaheuristics)
12
 Simulated annealing
13
 Genetic algorithms
14
 Artificial neural networks

Instructor(s)

Professor
Cüneyt F. Bazlamaçcı
Other MS Courses