The aim of the course is to describe and to analyse genetic algorithms and (other) evolution strategies in sufficient detail, such that the student is able to decide whether these methods for search and optimisation are suited to solve a particular problem, and how to choose the appropriate methods (e.g. selection of appropriate 'genetic operators').
Some model problems are studied and some case studies are discussed. Much attention is also paid to performance and implementation issues.
In the exercise sessions, the students analyse in detail some of the methods (e.g. 'genetic operators') and design a genetic algorithm for a particular problem.
For the project, the student use a matlab-based software package to perform a critical analysis of the performance, both in terms of quality of the result as in terms of the computational cost, of a genetic algorithm for a model problem (e.g. the traveling salesman problem. Implementation of some other genetic operators and the analysis of their performance is appreciated. A report on the results and their analysis must be submitted and will be discussed during the oral examination.
Basic (undergraduate) courses in informatics (programming, algorithms) and mathematics (analysis).
Articles and literature
Toledo / e-platform
Is also included in other courses
- Master in de statistiek (Biometrics) 120 ects.
- Master in de statistiek (General Statistical Methodology) 120 ects.
- Master in de ingenieurswetenschappen: wiskundige ingenieurstechnieken 120 ects.
- Master in de ingenieurswetenschappen: biomedische technologie 120 ects.
- Master of Artificial Intelligence 60 ects.
- Master of Statistics (Biometrics) 120 ects.
- Master of Statistics (General Statistical Methodology) 120 ects.
- Master in de informatica (uitdovend, enkel 2e fase) (Specialisation: Artificial Intelligence) 120 ects.
- Master in de ingenieurswetenschappen: computerwetenschappen (Specialisation: Artificial Intelligence) 120 ects.
- Master in de ingenieurswetenschappen: computerwetenschappen (Specialisation: Computational Informatics) 120 ects.
- Master of Engineering: Biomedical Engineering 120 ects.
- Introduction and situation of the course. The basic genetic algorithm.
- More details on genetic operators (selection, cross-over, mutation) and on the representation of the population for some model problems
- Theoretical foundation: schemata theorem; strategies to avoid premature convergence and to improve convergence or execution time: sampling mechanisms, adaptation of the fitness function, varying population size
- Handling constraints
- Related methods for search and optimisation: genetic programming, evolution strategies. Comparison with other methods for search and optimisation: gradient-based methods (hill climbing), simulated annealing, tabu search
- Analysing genetic algorithms using fitness landscape analysis
- Model problems: travelling salesman problem, transportation problem
- Case studies: e.g. timetabling, concept learning, classification, evolving 3D morphology and behaviour by competition
- Software packages
Oral exam (open book), partly project work.