Neural Networks and Evolution Methods (FSI-VSC-K)

Academic year 2018/2019
Supervisor: prof. Ing. Radomil Matoušek, Ph.D.  
Supervising institute: ÚAI all courses guaranted by this institute
Teaching language: Czech
Aims of the course unit:
The course objective is to make students familiar with basic resources of Soft Computing, potential and adequacy of their use in engineering problems solving.
Learning outcomes and competences:
Understanding of basic methods of Soft Computing and ability of their implementation.
Prerequisites:
The knowledge of basic relations of the optimization, statistics, graphs theory and programming.
Course contents:
The course introduces basic approaches to Soft Computing and classical methods used in the field. Practical use of the methods is demonstrated on solving simple engineering problems.
Teaching methods and criteria:
The course is taught through lectures explaining the basic principles and theory of the discipline. Exercises are focused on practical topics presented in lectures.
Assesment methods and criteria linked to learning outcomes:
Course-unit credit requirements: submitting a functional software project which uses implementation of selected AI method. Project is specified in the first seminar. Systematic checks and consultations are performed during the semester. Each student has to get through one test and complete all given tasks. Student can obtain 100 marks, 40 marks during seminars (20 for project and 20 for test; he needs at least 20), 60 marks during exam (he needs at least 30).
Controlled participation in lessons:
The attendance at lectures is recommended, at seminars it is obligatory. Education runs according to week schedules. The form of compensation of missed seminars is fully in the competence of a tutor.
Type of course unit:
    Guided consultation  1 × 17 hrs. optionally                  
    Controlled Self-study  1 × 35 hrs. compulsory                  
Course curriculum:
    Guided consultation 1. Introduction, Soft Computing concept explanation in Artificial Intelligence.
2. Architectures and classification of neural networks. Perceptron, ADALINE.
3. Feed-forward neural networks, single- and multilayer perceptron. Learning: error back-propagation as iterative minimisation of the mean quadratic error.
4. Cluster analysis, dimensionality reduction, Principal component analysis.
5. RBF and RCE neural networks. Topologic organized neural network (competetive learning, Kohonen maps).
6. Neural networks as associative memories (Hopfield networks, BAM), behaviour, state diagram, attractors, learning.
7. LVQ neural networks, ART neural networks
8. Fuzzy sets, fuzzy logic and fuzzy numbers, Fuzzy inference. ANFIS
9. Evolutionary algorithms (genetic algorithms, evolutionary strategy, grammatical evolution, genetic programming).
10. Selected metaheuristics for optimization (HC12, Simulated anealing).
11. Swarm intelligence (PSO, ACO, DE, SOMA)
12. Deterministic chaos.
13. Hybrid approaches and aplications (neural networks, fuzzy logic, genetic algorithms sets).
    Controlled Self-study Seminars related to the lectures in the previous week.
Literature - fundamental:
1. Aliev,R.A, Aliev,R.R.: Soft Computing and its Application, World Scientific Publishing Co. Pte. Ltd., 2001, ISBN 981-02-4700-1
2. Sima,J., Neruda,R.: Theoretical questions of neural networks, MATFYZPRESS, 1996, ISBN 80-85863-18-9
3. Munakata, T.: Fundamentals of the New Artificial Intelligence, Springer-Verlag New York, Inc., 1998. ISBN 0-387-98302-3
4. Švarc, I., Matoušek, R., Šeda, M., Vítečková, M.: Automatizace-Automatické řízení, skriptum VUT FSI v Brně, CERM 2011.
The study programmes with the given course:
Programme Study form Branch Spec. Final classification   Course-unit credits     Obligation     Level     Year     Semester  
M2I-K combined study M-AIŘ Applied Computer Science and Control P linked to branch B-AIR Cr,Ex 5 Compulsory 2 1 S
M2I-K combined study M-AIŘ Applied Computer Science and Control -- Cr,Ex 5 Compulsory 2 1 S