Courses – Fakultet tehničkih nauka u Novom Sadu

Subject: Heuristic Algorithms (17.EM503)

Native organizations units: Department of Power, Electronic and Telecommunication Engineering

General information:

Category	Theoretical-methodological
Scientific or art field	Electronics
ECTS	5

Most engineering problems of interest use complex algorithms, and use up considerable computer resources (time, space, the number of processors). When lacking the efficient deterministic and approximative algorithms for solving complex problems, adequately designed and applied (meta)heuristcs produce acceptable (suboptimal) solutions in an acceptable time frame. The educational objective of the course is to offer a well organized and comprehensive comparative overview of (meta)heuristics and soft-computing techniques widley used in practical engineering solutions of difficult algorithm problems.

Students who complete the course: -understand basic (meta)heuristics and soft-computing techniques for algorithm problem solving, -develop the ability to classify problems (to determine the level of algorithm difficulty of the problem, to reduce problem to the existing problem types), -can work with different programming libraries that use (meta)heuristics of general and specific applicaition

Types of algorithms: deterministic, approximative, randomized, heuristic and metaheuristic; why and when to use (meta)heuristics. Traditional deterministic searching methods. Simple heuristic methods: types of heuristics, heuristic design, local search heuristics, heuristics based on local search, interative local search. Metaheuristics: evolutionary computation (EC), evolutionary algorithms (EA), evolutionary strategies (ES), evolutionary programming (EP), genetic algorithms (GA), genetic programming (GP), hybrid methods; tabu search (TS), simulated annealing (SA), quantum annealing (QA), ant colony optimization (ACO), swarm intellience (SI), memetic algorithms (MA). Soft-computing: artificial neural networks (ANN), cell neural networks (CNN), fuzza logic based algorithms (FA), hybrid methods (neuro-fuzzy, fuzzy-genetic, etc.). The use of heuristics, metaheuristics and soft-computing in algorithm solutions to difficult (optimization) engineering problems, such as linear programming (LP), integral programming (IP), 0-1 integral programming (0-1 IP), non-linear programming (NLP), single objective (SO) and multi-objective (MO) optimization goals.

Lectures. Auditory practice. Computer practice. Laboratory practice. Tutorial work.

Authors	Title	Year	Publisher	Language
T. Back, David B. Fogel, Z. Michalewicz	Handbook of Evolutionary Computation	1997	Springer	English
Zbigniew Michalewicz, David B. Fogel	How to Solve It: Modern Heuristics	2004	2nd ed. Revised and Extended edition, Springer	English
J.-S. R. Jang, C.-T. Sun, E. Mizutani	Neuro-Fuzzy and Soft Computing	1996	Prentice-Hall	English
Daniel Ashlock	Evolutionary Computation for Modeling and Optimization	2006	Springer	English