Course details

Evolutionary Computation

EVD Acad. year 2012/2013 Summer semester

Current academic year

Evolutionary computation in the context of artificial intelligence and optimization problems with NP complexity. Paradigm of genetic algorithms, evolutionary strategy, genetic programming and another evolutionary heuristics. Theory and practice of standard evolutionary computation. Advanced evolutionary algorithms based on graphic probabilistic models (EDA - estimation of distribution algorithms). Parallel evolutionary algorithms. A survey of representative applications of evolutionary algorithms in multi-objection optimization problems, artificial intelligence, knowledge based systems and digital circuit design. Techniques of rapid prototyping of evolutionary algorithms.

Guarantor

Schwarz Josef, doc. Ing., CSc.

Language of instruction

Czech, English

Completion

Examination

Time span

  • 39 hrs lectures

Department

Department of Computer Systems (UPSY)

Subject specific learning outcomes and competences

Skills and approaches in solution of hard optimization problems.

Learning objectives

To inform the students about up to date algorithms for solution of complex, NP complete problems.

Prerequisite knowledge and skills

There are no prerequisites

Study literature

  • Fogel D., B.: Evolutionary computation: Toward a new philosophy of machine intelligence. IEEE Press, New York, 2000, ISBN 0-7803-5379-X.

Fundamental literature

  • Back, J: Evolutionary algorithms, theory and practice, New York, 1996.
  • Goldberg, D., E.: The Design of Innovation: Lessons from and for Competent Genetic Algorithms. Boston, MA: Kluwer Academic Publishers, 2002. ISBN: 1402070985.
  • Kvasnička V., Pospíchal J., Tiňo P.: Evoluční algoritmy. Vydavatelství STU Bratislava, 2000, str. 215, ISBN 80-227-1377-5.

Syllabus of lectures

  • Evolutionary algorithms, theoretical foundation, basic distribution.
  • Genetic algorithms (GA), schemata theory.
  • Advanced genetic algorithms
  • Repesentative combinatorial optimization problems.
  • Evolution strategies.
  • Genetic programming.
  • Advanced estimation distribution algorithms (EDA).
  • Variants of EDA algorithms, UMDA, BMDA and BOA.
  • Simulated annealing.
  • Methods for multicriterial and multimodal problems. Selection and population replacement.
  • Techniques for fast prototyping. Structure of development systems and GA library.
  • New evolutionary paradigm: immune systems,  differential evolution, SOMA.
  • Typical application tasks.

Progress assessment

Study evaluation is based on marks obtained for specified items. Minimimum number of marks to pass is 50.

Controlled instruction

Project defence, software project based on a variant of evolutionary algorithms or the  presentation of the assigned task.

Back to top