COURSE SYLLABUS

IE 420      HEURISTIC METHODS FOR OPTIMIZATION      (3-0) 3

URL: http://ie.atilim.edu.tr/~ie420

Catalog Data:

Introduction of a variety of important, main-stream heuristic techniques, both traditional and modern, for solving combinatorial problems. Reasons for the existence of heuristic techniques, their applicability and capabilities.

Text Book:

Reeves, C. R., Modern Heuristic Techniques for Combinatorial Problems, John Wiley & Sons, 1993.

References:

·         Sait, S.M., and Youssef, H., Iterative Algorithms with Applications in Engineering, IEEE Press, 1999. 

·         Papadimitriou, C.H., and Steiglitz, K., Combinatorial Optimization: Algorithms and Complexity, Prentice-Hall, 1982.

·         Nemhauser, G.L., and Wolsey, L.A., Integer and Combinatorial Optimization, John Wiley & Sons, 1998.

·         Lawler, E.L., Lenstra, J.K., Rinnooy Kan, A.H.G., and Shmoys, D.B., The Traveling Salesman Problem, John Wiley & Sons, 1985.

Prerequisites by Topic:

Basics of optimization theory; computer programming.

Method for Assessing Student Knowledge of Prerequisite Topics:

Pre-requisite exam will be given on the prerequsite topics.

Goals:

Upon successful completion of this course, students should gain knowledge of how and why heuristic techniques work, when they should be applied and their relative merits with respect to each other and with respect to more traditional approaches, such as mathematical programming.

Objectives:

·         To help students gain insight about the newer and most common heuristic search methods

·    To help students analyze some common heuristics, such as simulated annealing, genetic algorithms, evolutionary strategies and tabu search

·       To provide student knowledge with some other heuristic methods, such as neural networks and random methods

Topics:

1.        Introduction: computational growth rate, algorithmic complexity and combinatorial problems 1 week)

2.     Branch-and-Bound: branching, bounding, node development, dominance, relaxation to provide bounds and integer programming (2 weeks)

3.     Lagrangian relaxation (2 weeks)

4.     Local search: neighbourhoods, local and global optimality, constructive and improvement heuristic techniques (2 weeks)

5.     Simulated annealing: general approach, cooling schedules and variants (1 week)

6.     Genetic algorithms: populations, reproduction, crossover, mutation, demes, competition and genetic programming (2 weeks)

7.     Tabu search: short term memory, tabu status, aspiration, intensification and diversification (2 weeks)

8.     Other methods and techniques: neural networks, random methods, hybrid methods, Great Deluge algorithm, record-to-record transfer and parallel implementation (2 weeks)

Computer Usage:

A term project and homework assignments which require coding will be given.

Laboratory Projects:

A project that requires implementation of one or more heuristic algorithms .

Contribution to Professional Component:

1.        Mathematics and Basic Science                 0 credits

2.       Engineering Science or Design                   3 credits

3.       General Education                                     0 credits