Bio-inspired Computing (FSI-VBC-K)

Academic year 2021/2022
Supervisor: prof. Ing. Radomil Matoušek, Ph.D.  
Supervising institute: ÚAI all courses guaranted by this institute
Teaching language: Czech
Aims of the course unit:
Goal of the course is to introduce students to modern tools of biology inspired computing and options and appropriate usage for solving engineering tasks.
Learning outcomes and competences:
Knowledge: Students will know basic principles and algorithms of presented methods usable in continuous and combinatorial optimization and their options, restrictions and potential for implementation.
Skills: Ability to use these methods to solve practical engineering problems where methods of mathematical optimization may not provide acceptable results.
Prerequisites:
Statistics and Optimization Methods I.
Course contents:
The course introduces basic and advanced methods of so called biology inspired computing. Focus is on practical implementation of this special class of artificial intelligence algorithms. Usability of the methods is demonstrated with mathematical and engineering problems.
Teaching methods and criteria:
The course is taught in the form of lectures, which have the character of an explanation of the basic principles and theory of the discipline, incl. presentation of practical applications. The exercise is focused on the practical mastery of the material covered in lectures in the form of team projects. Due to the circumstances, the subject can be taught remotely.
.
Assesment methods and criteria linked to learning outcomes:
Requirements for credit: Students will be divided into teams. They must submit 4 functioning software projects for each team. Each team member must be able to present and understand the projects. Concrete specification will be on the laboratory exercise. Consultations are provided and project progress is checked continuously. Individual projects are in completion. Maximum points form exercises is 100, credit limit is 60.
Controlled participation in lessons:
Attendance at seminars is controlled. An absence can be compensated for via solving additional problems.
Type of course unit:
    Guided consultation in combined form of studies  1 × 13 hrs. compulsory                  
    Guided consultation  1 × 26 hrs. optionally                  
Course curriculum:
    Guided consultation in combined form of studies Teaching will be divided into 4 blocks reflecting real usage of biology inspired computation. Students will work in groups and compare in competition the obtained results.
A. Implementation of GA and solution of concrete optimization task*
B. Implementation of chosen meta-heuristics and solution of concrete optimization task *
C. Implementation of CGA for evolutionary design of hardware
D. Implementation of Cellular automata
*Tasks of combinatorial, integer and mixed optimization (TSP, QAP, controller design, symbolic regression, global optimization of multi-modal functions, etc.)
    Guided consultation B1: Biology inspired computation - introduction. History and division of evolutionary computing techniques (ECT). Standard genetic algorithms (SGA). Holland's schema theorem. Building Block Hypothesis.
B2: Advanced GA. Problem coding methods. Combinatorial optimization using GA. 4. Grammar Evolution (GE). Genetic Programming (GP). Symbolic regression tasks. Cartesial Genetic Programming (CGA). Evolutionary design of combinational logic circuits.
B3: Evolution Strategy (ES). Differential Evolution (DE). Representation. Basic models. Binary string searching algorithm HC12. Nelder-Mead algorithm. Algorithms using patterns. Bayesian optimization algorithms.
B4: Swarm algorithms I. (Ant Colony strategy, Bee Colony Optimization). Swarm algorithms II. (Particle Swarm Optimization, Firefly algorithm, SOMA).
B5: Cellular automata I – theory basics. Cellular automata II – practical applications.
B6: Summary – colloquium.
Literature - fundamental:
1. KVASNIČKA, Vladimír. Evolučné algoritmy. Bratislava: Vydavateľstvo STU, 2000. Edícia vysokoškolských učebníc. ISBN isbn80-227-1377-5.
2. ZELINKA, Ivan. a kol. Evoluční výpočetní techniky. Principy a aplikace. Praha, BEN 2009.
Literature - recommended:
1. HAUPT, R. L., HAUPT, S. E. Practical Genetic Algorithms. John Wiley & Sons 1998.
2. DORIGO, M., STüTZLE, T. Ant Colony Optimization. MIT Press 2004.
The study programmes with the given course:
Programme Study form Branch Spec. Final classification   Course-unit credits     Obligation     Level     Year     Semester  
N-AIŘ-K combined study --- no specialisation -- GCr 4 Compulsory 2 2 W