|
|
|
||
|
Models of evolution, genetic algorithms, representation and operators of selection, mutation and crossover. Problem solving by means of evolutionary computation. Theoretical properties of simple genetic algorithm. Schemata theorem and building block hypothesis, probabilistic models. Evolutionarz machine learning, Michigan vs. Pittsburg approach, classifier systems.
Last update: Neruda Roman, Mgr., CSc. (02.05.2006)
|
|
||
|
To learn basic techniques used in evolutionary algorithms. Show connections with related topics of datamining and learning. Last update: Hric Jan, RNDr. (07.06.2019)
|
|
||
|
In order to pass the course, the student must obtain the credit for the seminar and pass an exam. The credit is given for solving assignments from the seminar. The nature of study verification excludes the possibility of its repetition. Last update: Pilát Martin, doc. Mgr., Ph.D. (13.10.2017)
|
|
||
|
Mitchell, M.: Introduction to genetic algorithms. MIT Press, 1996.
Goldberg, D.: Genetic algorithms in search optimization and machine learning, Addison-Wesley, 1989.
Holland, J.: Adaptation in natural and artificial systems, MIT Press, 1992 (2nd ed).
Holland, J.: Hidden order, Addison-Wesley, 1995. Last update: Pilát Martin, doc. Mgr., Ph.D. (04.11.2019)
|
|
||
|
The exam is oral with time for written preparation. The requirements correspond to the syllabus in the extent presented during the lectures. A part of the exam asks to design an evolutionary algorithm for a given problem. Last update: Pilát Martin, doc. Mgr., Ph.D. (13.10.2017)
|
|
||
|
Models of evolution, basic approaches and notions. Population, recombination, fitness evaluation.
Genetic algorithms, solution encoding in a chromozome, basic operators of selection, mutation, crossover.
Selection, objective function, dynamic vs. static, roulette-wheel selection, tournaments, elitism.
Schema theorem, building block hypotheses, implicit paralallelism.
Probabilistic models of simple genetic algorithm, finite and infinite population.
Machine learning and data mining, evoluion of expert systems, internal representation, Michigan vs. Pittsburg approach.
Clasifier systems, if-then rules, bucket brigade algorithm, Q-learning, production systems. Last update: Neruda Roman, Mgr., CSc. (02.05.2006)
|